Analyses to Inform the Potential Use of Standardized ......Beginning in October 2016, standardized patient assessment data elements that measure functional status were introduced to

March 2019

Analyses to Inform the Use of Standardized Patient Assessment

Data Elements in the Inpatient Rehabilitation Facility Prospective

Payment System

Report

Prepared for

Center for Medicare and Medicaid Services (CMS)

Assistant Secretary for Planning and Evaluation (ASPE)

Prepared by

Melissa Morley, PhD Benjamin Silver, PhD

Anne Deutsch, RN, PhD, CRRN Nicole Coomer, PhD Allison Dorneo, BA

Laura Coots Daras, PhD Melvin Ingber, PhD

RTI International

3040 E. Cornwallis Road Research Triangle Park, NC 27709

RTI Project Number 0212050.030

The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

_________________________________ RTI International is a registered trademark and a trade name of Research Triangle Institute. The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

ANALYSES TO INFORM THE USE OF STANDARDIZED PATIENT ASSESSMENT DATA ELEMENTS IN THE INPATIENT REHABILITATION FACILITY PROSPECTIVE

PAYMENT SYSTEM

by Melissa Morley, Project Director Benjamin Silver, Anne Deutsch, Nicole Coomer, Allison Dorneo, Laura Coots Daras, and

Melvin Ingber

RTI International

Contract No. HHSP23320095651WC, Order No. HHSP23337033T

March 2019

This project was funded by the Centers for Medicare & Medicaid Services and Assistant Secretary for Planning and Evaluation under Contract No. HHSP23320095651WC, Order No. HHSP23337033T. The statements contained in this report are solely those of the authors and do not necessarily reflect the views or policies of the Centers for Medicare & Medicaid Services or the Assistant Secretary for Planning and Evaluation. RTI assumes responsibility for the accuracy and completeness of the information contained in this report.

iii The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

Contents

Section Page

Executive Summary ES-1

Section 1 Background 1-1 1.1 Introduction ............................................................................................ 1-1 1.2 Overview of the IRF PPS ........................................................................... 1-2 1.3 Use of Assessment Data in the IRF PPS: Motor Function and Cognitive

Function ................................................................................................. 1-3 1.4 Standardized Patient Assessment Data Elements .......................................... 1-4

Section 2 Analytic Approach 2-1 2.1 Data and Sample ..................................................................................... 2-1 2.2 Classification and Regression Tree Analysis: Using Standardized Patient

Assessment Data Elements to Generate CMGs ............................................. 2-1 2.3 Motor Score Using Standardized Patient Assessment Data Elements ............... 2-2 2.4 Cognitive Function Using Standardized Patient Assessment Data Elements ...... 2-5 2.5 Costs ...................................................................................................... 2-6 2.6 Results of CART Analysis ........................................................................... 2-6 2.7 Payment Weight Calculations for Standardized Patient Data Element–

Based CMGs .......................................................................................... 2-14

References R-1

Appendixes

A IRF Patient Assessment Instrument ............................................................ A-1 B FY 2019 IRF PPS CMGs and Payment Weights .............................................. B-1 C Standardized Patient Assessment Data Elements and Functional

Independence Measure (FIM™) Item Descriptives ........................................ C-1 D Correlation of Weighted Versus Unweighted Motor Score .............................. D-1 E Cross Validated R-Squared by RIC .............................................................. E-1

iv The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

Figure

Number Page

1. RICs in the IRF PPS ....................................................................................... 1-2

v The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

Tables

Number Page

1. IRF-PAI FIM™ Items Used in IRF PPS Motor Score ............................................. 1-3 2. FIM™ Levels ................................................................................................. 1-4 3. IRF-PAI FIM™ Items Used in the IRF PPS Cognitive Score ................................... 1-4 4a. IRF-PAI Standardized Patient Assessment Data Elements: Motor Score ................ 1-6 4b. IRF-PAI Standardized Patient Assessment Data Elements: Cognitive Function ....... 1-6 5. Self-Care and Mobility Rating Scale for Standardized Patient Assessment Data

Elements ...................................................................................................... 1-7 6. Recoding for IRF-PAI Item H0350: Bladder Continence ...................................... 2-4 7. Recoding for IRF-PAI Item H0400: Bowel Continence ......................................... 2-4 8. Memory Using BIMS and Staff Assessment of Mental Status ............................... 2-6 9. CMG Definitions Using Standardized Patient Assessment Data Elements ............... 2-9 10. Comparison of RIC-Level Average Payment Weights ........................................ 2-17

vi The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

[This page intentionally left blank]

ES-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

EXECUTIVE SUMMARY

The purpose of this report is to provide analyses to inform the use of standardized patient assessment data elements collected on admission in the Inpatient Rehabilitation Facility Prospective Payment System (IRF PPS). The report summarizes the use of assessment data in the current IRF PPS, describes the process used to substitute standardized patient assessment data elements collected on admission into the IRF PPS, and presents the case-mix groups (CMGs) and payment weights based on those elements. The results presented here are based on the analysis of Fiscal Year (FY) 2017 and FY 2018 data and represent an update and refinement to earlier work using only FY 2017 data (Morley, Silver, Deutsch, & Ingber, 2018).

The analyses were conducted under the assumption that all other aspects of the inpatient rehabilitation facility (IRF) payment system remain unchanged, including the rehabilitation impairment category (RIC) structure, the assignment of comorbidity tiers, and the methodology for calculating the payment weights. The focus of this work was ensuring that the CMGs within RICs accurately reflect patient costs when using standardized patient assessment data elements collected on admission in place of the Functional Independence Measure (FIM™) items.

The data used in these analyses were drawn from the FY 2017 and FY 2018 Medicare Inpatient National Claims History and Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI) data files. Consistent with the approach used in the development of the current IRF PPS (Carter et al., 2002), RTI International used Classification and Regression Tree (CART) analysis to develop CMGs using standardized patient assessment data elements collected on admission, including motor function, cognitive function, and age. A refinement to the work presented here includes the use of a weighted motor function score, consistent with the current IRF PPS.

The CART models using the standardized patient assessment data elements collected on admission yielded 97 CMGs (including RICs 50 and 51, which remain unchanged). There are 92 CMGs in the FY 2019 IRF PPS. Although the overall number of CMGs is similar, there are small changes in the number of CMGs per RIC. For example, RIC 1 contains 10 CMGs in the FY 2019 IRF PPS, but 7 when using the standardized patient assessment data elements. Motor score emerged as the key function variable in the definition of the CMGs across all RICs. Cognitive function is not used to define the CMGs emerging from the use of the standardized patient assessment data elements, though it was included in the CART modeling.

Analyses to Inform the Use of Standardized Patient Assessment Data Elements in the Inpatient Rehabilitation Facility Prospective Payment System

ES-2 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.


1-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

SECTION 1 BACKGROUND

1.1 Introduction

The purpose of this report is to provide analyses to inform the use of standardized patient assessment data elements collected on admission in the Inpatient Rehabilitation Facility Prospective Payment System (IRF PPS). The results presented here are based on the analysis of Fiscal Year (FY) 2017 and FY 2018 data and represent an update and refinement to earlier work using FY 2017 data only (Morley, Silver, Deutsch, and Ingber, 2018).

The IRF PPS, implemented in 2002, is based on Functional Independence Measure (FIM™) items collected in the Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI). In the IRF PPS, patients are assigned to one of 87 case-mix groups (CMGs) on the basis of diagnosis requiring rehabilitation, motor function, cognitive function, and age. An additional five CMGs are used if the patient either dies or is discharged in 3 days or less.

Beginning in October 2016, standardized patient assessment data elements that measure functional status were introduced to the IRF-PAI. Collection of these items began as part of the IRF Quality Reporting Program (QRP). The standardized patient assessment data elements were also introduced to the Minimum Data Set, which is the assessment instrument used for skilled nursing facilities, and the Long-Term Care Hospital Continuity Assessment Record and Evaluation (CARE) data set, which is the assessment instrument used for long-term care hospitals. Standardized patient assessment data elements were introduced to the Outcome and Assessment Information Set (OASIS), the assessment instrument used for home health agencies, beginning in January 2019. The IRF Quality Reporting Program includes four functional outcome measures related to self-care and mobility that are based on standardized patient assessment data elements. These quality measures were finalized in the FY 2016 IRF PPS Final Rule (80 FR 47111 through 47117). Since October 2016, IRF providers have collected both standardized patient assessment data elements and FIM™ items (along with function modifier items that are used to generate FIM™ items). In the FY 2019 IRF PPS Final Rule, CMS finalized the removal of the FIM™ instrument from the IRF-PAI beginning in FY 2020 to reduce burden by eliminating the overlap between the FIM™ items and the standardized patient assessment data elements. This also supports CMS’s broader goal of standardizing data collection across PAC settings.

The next sections of this report summarize the use of assessment data in the IRF PPS, describe the process RTI International used to substitute standardized patient assessment data elements collected on admission into the IRF PPS, and present the CMGs and payment weights based on the use of the standardized patient assessment data elements collected on admission.



1.2 Overview of the IRF PPS

Under the IRF PPS, Medicare fee-for-service payments are made to IRFs on a per-discharge basis, and patients are assigned to one of 92 CMGs. Of the 92 CMGs, 87 are assigned based on diagnoses requiring rehabilitation, motor function, cognitive function, and age. Five CMGs assignments are based on short-stay status or death during the IRF stay. Data on motor function and cognitive function are obtained from IRF-PAI assessments collected by all IRF providers at admission. The Fiscal Year 2017 IRF-PAI can be found in Appendix A.

Each of the diagnosis-based rehabilitation impairment categories (RICs), derived from the admission impairment group code on the IRF-PAI, has a RIC-specific set of CMGs based on the characteristics of the patients in that RIC. The number of CMGs can vary by RIC. For example, in the current IRF PPS, there are 10 CMGs for RIC 1, “stroke,” but only 4 CMGs for RIC 14, “cardiac.” The CMGs reflect differences in costs by different levels of motor function, cognitive function, and age. Figure 1 shows the 23 RICs in the IRF PPS.

Figure 1. RICs in the IRF PPS

1. Stroke 2. Traumatic brain injury 3. Non-traumatic brain injury 4. Traumatic spinal cord injury 5. Non-traumatic spinal cord injury 6. Neurological 7. Fracture of lower extremity 8. Replacement of lower extremity 9. Other orthopedic 10. Amputation, lower extremity 11. Amputation, non-lower extremity 12. Osteoarthritis 13. Rheumatoid, other arthritis

14. Cardiac 15. Pulmonary 16. Pain syndrome 17. Major multiple trauma without brain or

spinal cord injury 18. Major multiple trauma with brain or spinal

cord injury 19. Guillain-Barré 20. Miscellaneous 21. Burns 50. Short stay 51. Expired

Payment weights are based on a combination of CMG and comorbidity tier using an established comorbidity list applicable to all CMGs (with the exception of RIC 50, “short stay,” and RIC 51, “expired”). There are four comorbidity tiers, each reflecting an increasing level of severity. Each year, CMS updates a national base payment amount (called a standard payment conversion factor) and payment weights for each CMG and comorbidity tier combination. The IRF PPS CMGs and relative weights for Fiscal Year 2019 are shown in Appendix B. To derive payment, the standard payment conversion factor is multiplied by the payment weight associated with each CMG. Payments are also adjusted for geographic differences in wages, by the proportion of each facility’s care furnished to low-income individuals, by rural status, and by teaching status, as applicable. Finally, patients who are



transferred to another inpatient setting after a below-average length of stay in the IRF (specific to each CMG) are paid for on a per-diem basis.

1.3 Use of Assessment Data in the IRF PPS: Motor Function and Cognitive Function

The IRF PPS uses admission FIM™ items from the IRF-PAI to construct a motor score and a cognitive score. These scores in turn are used for CMG assignment. The FIM™ items used to create the motor score are shown in Table 1. The IRF PPS uses a weighted motor score that was developed as part of the initial payment system development work by RAND and CMS (Carter et al., 2002). Rather than applying an equal weight to each FIM™ item to generate a motor score, items are weighted to reflect their relative contribution to costs of care. The weights associated with each item in constructing the motor score are reported in Table 1. Table 2 shows the rating scale for the FIM™ Items. The rating scale reflects a patient’s need for assistance and differentiates between total dependence and complete independence. Activities that did not occur, originally coded to “0,” are recoded to the most dependent level on the rating scale (Level 1, “Total Assistance”), except for toilet transfer, where ”0” is recoded to Level 2, “Maximal Assistance.” The range for the motor function score is 12–84 (12 items assessed on a scale of 1–7), with higher scores indicating higher ability.

Table 1. IRF-PAI FIM™ Items Used in IRF PPS Motor Score

Item Number Weight

Eating 39Aa 0.6 Grooming 39Ba 0.2 Bathing 39Ca 0.9 Dressing—upper 39Da 0.2 Dressing—lower 39Ea 1.4 Toileting 39Fa 1.2 Bladder 39Ga 0.5 Bowel 39Ha 0.2 Bed, chair, wheelchair transfer 39Ia 2.2 Toilet transfer 39Ja 1.4 Walk/wheelchair 39La 1.6 Stairs 39Ma 1.6

SOURCE: IRF-PAI.



Table 2. FIM™ Levels

Level

No Helper

7 Complete independence (timely, safely)

6 Modified independence (device)

Helper—Modified Dependence

5 Supervision (subject = 100%)

4 Minimal assistance (subject = 75% or more)

3 Moderate assistance (subject = 50% or more)

Helper—Complete Dependence

2 Maximal assistance (subject = 25% or more)

1 Total assistance (subject less than 25%)

0 Activity does not occur; use this code only at admission

SOURCE: IRF-PAI.

Five cognitive function items based on FIM™ are included in the cognitive score (Table 3). To calculate the cognitive score for payment, these items are summed (with equal weighting). The range for the cognitive score is 5–35 (5 items assessed on a scale of 1–7), with higher scores indicating higher ability. The cognitive items use the same rating scale as the motor function items.

Table 3. IRF-PAI FIM™ Items Used in the IRF PPS Cognitive Score

Item Number

Comprehension 39N

Expression 39O

Social interaction 39P

Problem solving 39Q

Memory 39R

SOURCE: IRF-PAI.

1.4 Standardized Patient Assessment Data Elements

Beginning in October 2016, Medicare required IRFs to complete standardized patient assessment data elements for Hearing, Speech, and Vision (Section B); Cognitive Patterns (Section C); Functional Abilities and Goals – Self-Care and Mobility (Section GG); and Bladder and Bowel (Section H) on the IRF-PAI. Though the content of the FIM™ items (and



the function modifier items that are used to generate the FIM™ items) overlaps with the standardized patient assessment data elements (e.g., eating, dressing, transfer), the items differ in the specific item definitions and the rating scale used for scoring. Standardized patient assessment data elements and FIM™ items (and the function modifier items that are used to generate the FIM™ items) are both collected within a 3-day period from a patient’s admission, but the instructions for assessing patient performance differ. The standardized patient assessment data elements assess a patient’s usual performance during the assessment period, in contrast to FIM™ items (and the function modifier items that are used to generate the FIM™ items), which assess a patient’s most-dependent status (i.e., lowest score) during the assessment period. The self-care and mobility standardized patient assessment data elements use a six-level rating scale, whereas FIM™ (and the function modifiers that are used to generate the FIM™ items) uses a seven-level scale.

Tables 4a and 4b outline the standardized patient assessment data elements used in RTI’s analysis of the IRF PPS, and Table 5 shows the rating scale for the self-care and mobility data elements. RTI considered all self-care and mobility items in Section GG of the IRF-PAI for inclusion in the motor score and came to the set shown in Table 4a after considering both item multicollinearity and activities attempted at admission. Walking on uneven surfaces, car transfer, steps, and pick up object were not included in the motor score because these activities are less likely to be attempted on admission, as the patient’s medical condition or safety concerns may prevent performance of the activity. See Appendix C for descriptive statistics on assessment items. Although roll left and right was included in the analyses presented in the 2018 report (Morley et al., 2018), the results of additional analyses identified a high degree of multicollinearity with other standardized patient assessment data elements. This item was inversely correlated with costs after controlling for each of the other self-care and mobility items. Note that all available standardized patient assessment data elements related to cognition were included in the analysis.

Given the differences in the item definitions and rating scales, using the standardized patient assessment data elements in place of FIM™ items requires more than a simple substitution into the current IRF PPS for the purposes of assigning patients to payment groups and computing payments. To incorporate the standardized patient assessment data elements into the payment system analysis, RTI considered the range of available items to construct a motor score and to account for cognition. A single motor score, rather than separate self-care and mobility scores, was used to be consistent with the current IRF PPS and because the analyses are conducted within RICs, that is, primary rehabilitation categories. Differentiating between self-care and mobility scores can be valuable when analyses are conducted in aggregate across diagnoses groups rather than by diagnosis group.



Table 4a. IRF-PAI Standardized Patient Assessment Data Elements: Motor Score

Item Number Item Description Weight (Total =18)

GG0130A1 Eating 2.7

GG0130B1 Oral hygiene 0.3

GG0130C1 Toileting hygiene 2.0

GG0130E1 Shower/bathe self 0.7

GG0130F1 Upper-body dressing 0.5

GG0130G1 Lower-body dressing 1.0

GG0130H1 Putting on/taking off footwear 1.0

GG0170B1 Sit to lying 0.1

GG0170C1 Lying to sitting on side of bed 0.1

GG0170D1 Sit to stand 1.1

GG0170E1 Chair/bed-to-chair transfer 1.1

GG0170F1 Toilet transfer 1.6

GG0170I1 Walk 10 feet 0.8

GG0170J1 Walk 50 feet with two turns 0.8

GG0170K1 Walk 150 feet 0.8

GG0170M1 One-step curb 1.4

H0350 Bladder continence 1.3

H0400 Bowel continence 0.7

Table 4b. IRF-PAI Standardized Patient Assessment Data Elements: Cognitive Function

Item Number Item Description

BB0700 Expression of ideas and wants

BB0800 Understanding verbal content

C0500 Brief Interview for Mental Status (BIMS) summary score

C0900 Memory/Recall (Staff Assessment)



Table 5. Self-Care and Mobility Rating Scale for Standardized Patient Assessment Data Elements

Self-Care and Mobility Rating Scale

Activities may be completed with or without assistive devices

06. Independent

05. Setup or clean-up assistance

04. Supervision or touching assistance

03. Partial/moderate assistance

02. Substantial/maximal assistance

01. Dependent

If activity was not attempted, code reason

07. Patient refused

09. Not applicable

88. Not attempted because of medical condition or safety concerns

Source: IRF-PAI

RTI refined the analyses of the FY 2017 and FY 2018 data by incorporating a weighted motor score. A similar approach is part of the current IRF PPS. As part of calculating the weights, RTI conducted analyses to assess the degree of multicollinearity between the standardized patient assessment data elements. RTI identified three pairs of items that are highly correlated: (1) “lower-body dressing” and “putting on/taking off footwear,” (2) “sit to lying” and “lying to sitting on side of bed,” and (3) “sit to stand” and “chair/bed-to-chair transfer.” Weights were calculated for these item pairs (using the average of each pair of item scores), and then each individual item in the pair was assigned half the calculated weight. Similarly, the weight calculation for walking was based on the item “walk 10 feet,” and the resulting weight was divided equally across all three walking items.

Note that wheelchair mobility items were not included as separate items in the motor score. The walking items, rather than a combination of the walking and wheelchair items, were used to measure mobility. Patients who do not walk are assigned to the most dependent response category for the walking items to reflect the greater resource use associated with patients who cannot walk. If wheelchair mobility item scores were included, then some wheelchair users would have higher motor scores because they were not completely dependent in wheelchair mobility. By using the walking items only to assess mobility, the motor score reflects increased resource use and need for assistance among patients who do not walk.



RTI generated a set of CMGs based on the standardized patient assessment data elements collected on admission to reflect the differences in items, definitions, and rating scales. Section 2 describes the data used and the complete analytic approach.


SECTION 2 ANALYTIC APPROACH

This section outlines the overall analytic approach for generating case-mix groups (CMGs) and corresponding payment weights using standardized patient assessment data elements collected on admission. The analyses were conducted under the assumption that all other aspects of the inpatient rehabilitation facility (IRF) payment system would remain unchanged, including the rehabilitation impairment category (RIC) structure, the assignment of comorbidity tiers, and the methodology for calculating the payment weights. The only focus of this work was ensuring that the CMGs within RICs would accurately reflect patient costs when using standardized patient assessment data elements collected on admission in place of the Functional Independence Measure (FIM™) items.

2.1 Data and Sample

The data used in these analyses were drawn from the Fiscal Year (FY) 2017 and FY 2018 Medicare Inpatient National Claims History and Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI) data files. Medicare Cost Report data were used to construct IRF stay-level costs.

The analytic sample included 753,429 observations of IRF claims and matching assessment data (claims data were pulled in January 2019). Assessments were matched to IRF claims where Medicare Part A was the primary payer for the stay and the Medicare Part A payment amount was greater than zero. Claims were matched to IRF-PAI assessments by beneficiary ID (Health Insurance Claim [HIC] number) and admission date. IRF-PAI assessment item 20A (Payment Source) was used to confirm the primary payer.

IRF stays occurring at critical access hospitals were excluded. Appendix C contains descriptive statistics on this sample for all IRF-PAI assessments items considered in the current analyses.

2.2 Classification and Regression Tree Analysis: Using Standardized Patient Assessment Data Elements to Generate CMGs

Consistent with the approach used in the development of the current IRF PPS (Carter et al., 2002), RTI International used Classification and Regression Tree (CART) analysis to develop CMGs using standardized patient assessment data elements collected on admission, including motor score, cognitive score, and age. CART uses a stepwise process in which the data are split into “nodes,” or groups based on an outcome of interest. In this case, data were grouped using IRF claim costs by identifying the covariate and cut-point at each split that contribute most to the model fit (Morgan, 2014).



The CART approach continues to split the sample until the contribution to model fit of any further splitting falls below a user-determined threshold. RTI set CART parameters to require the following:

• Nodes are no smaller than 100 stays.

• Additional splits yield an 0.5 percentage point increase in explanatory power.

This process results in a “tree” of rules that can be used to assign cases to CMGs.

RTI used a subset of the analytic sample in the CART analysis. The subset is meant to reflect the most-typical cases and therefore excluded beneficiaries who died during the IRF stay, those whose stay lasted 3 days or fewer, and those who were transferred to another inpatient setting. As previously noted, decedents and short stays have separate CMGs under the IRF PPS, whereas transfers with a below-average length of stay are reimbursed using per diem rates. Stays with excessively high cost amounts (where cost is more than 3 standard deviations from the mean) were also excluded so that a small number of very high costs cases would not drive the results. Ultimately, RTI used a sample of 551,503 IRF stays to generate the set of CMGs using standardized patient assessment data elements.

The independent variables used in the CART regression included weighted motor score, cognitive function (specifically memory and communication), and age. The specification of these variables based on the standardized patient assessment data elements is described below. The dependent variable in the CART regression was IRF stay costs. The construction of the cost variable is also described in the following sections. CART models were run separately for each RIC, using the same methodology that was used when the IRF payment system was initially developed. This allows for diagnosis-specific splits on weighted motor score, cognitive function, and age to reflect the characteristics and resource use of the patients within each RIC.

2.3 Motor Score Using Standardized Patient Assessment Data Elements

RTI constructed a weighted motor score using the standardized patient assessment data elements collected on admission presented in Table 4a (Section 1). RTI considered all self-care and mobility items in Section GG of the IRF-PAI for inclusion in the motor score and came to the set shown in Table 4a after considering both item multicollinearity and activities attempted at admission. Walking on uneven surfaces, car transfer, steps, and pick up object were not included in the motor score because these activities are less likely to be attempted on admission, as the patient’s medical condition or safety concerns may prevent performance of the activity. The frequency with which these items are not attempted on admission in the IRF decreases their relevance for predicting costs (see Appendix C). Although roll left and right was included in the analyses presented in the 2018 report



(Morley et al., 2018), the results of additional analyses identified a high degree of multicollinearity with other standardized patient assessment data elements. This item was inversely correlated with costs after controlling for each of the other self-care and mobility items.

Each of the motor function items among the standardized patient assessment data elements is scored on a six-point scale, with higher scores indicating greater independence (Table 5). Additional codes can be used to indicate why the activity was not attempted. If one of these “activity not attempted” codes was recorded or a dash was entered or a caret because of a skip pattern, the score for that item was recoded to 1 (Dependent) for all items except GG0170F (Toilet Transfer), which was recoded to 2 (Substantial/Maximal Assistance). This recoding approach is consistent with the current recoding approach in the IRF PPS.

The standardized data elements for bladder (H0350) and bowel (H0400) continence were included in the motor score to be consistent with the current Functional Independence Measure (FIM™) motor score used in the IRF payment system. Because the higher response codes reflect more impairment and higher resource use would be associated with higher codes (i.e., 0 = always continent, 4 = always incontinent), scores on the bowel and bladder items were reversed for inclusion in the motor score calculation. Scores were also adjusted so that the minimum score was 1 for bowel and bladder items to be consistent with the minimum score for other items in the motor score.

The item recoding approach for the bowel and bladder items differed from the approach used for the self-care and mobility items. This recoding approach was informed by clinical review and consultation with the IRF-PAI Training Manual. See Table 6 for a summary of the recoding for H0350 (Bladder Continence) and Table 7 for a summary of the recoding for H0400 (Bowel Continence). Cases coded as 9 (not applicable, e.g., indwelling catheter) were recoded to “always incontinent” for the bladder item and “frequently incontinent” for the bowel item, and missing codes (i.e., dash use) were recoded to “always continent” for both items. Cases coded as “no urine output” (e.g., renal failure) on the bladder item were recoded to “always continent.” This recoding is consistent with the current approach, because patients who do not void are coded as 7, “complete independence,” for the FIM Bladder item. We also note that use of renal dialysis is accounted for in the comorbidity tiers. Finally, on the bladder item, “always continent” and “stress incontinence only” were combined to a single level representing no incontinence.



Table 6. Recoding for IRF-PAI Item H0350: Bladder Continence

N

Recode for Use in Motor Score

4— Always

Continent

3— Incontinent Less Than

Daily

2— Incontinent

Daily

1— Always

Incontinent

H0350—Bladder Continence

0—Always continent X

1—Stress incontinence only X

2—Incontinent less than daily X

3—Incontinent daily X

4—Always incontinent X

5—No urine output X

9—Not applicable X

Missing X

Table 7. Recoding for IRF-PAI Item H0400: Bowel Continence

N

Recoded for Use in Motor Score

4—Always Continent

3—Occasionally Incontinent

2—Frequently Incontinent

1—Always Incontinent

H0400—Bowel Continence

0—Always continent X

1—Occasionally incontinent X

2—Frequently incontinent X

3—Always incontinent X

9—Not rated X

Missing X

The motor score used in the CART model was calculated using a weighted sum of the scores for each of the 18 items in Table 4a. The range of scores for the motor score is 18 to 104, with a higher score indicating higher functional status and greater level of independence.

The calculation of the weighted motor score is a refinement to the approach presented in the 2018 report (Morley et al., 2018). A similar weighted motor score calculation is part of



the current IRF PPS. As part of calculating the weights, RTI conducted analyses to assess the degree of multicollinearity between the standardized patient assessment data elements items and identified three pairs of items that are very highly correlated: (1) “lower-body dressing” and “putting on/taking off footwear,” (2) “sit to lying” and “lying to sitting on side of bed,” and (3) “sit to stand” and “chair/bed-to-chair transfer.” Weights were calculated for these item pairs, and then each item in the pair was assigned half the calculated weight. Similarly, the weight calculation for walking was based on the item “walk 10 feet,” and the resulting weight was divided equally across all three walking items. Note that wheelchair mobility items were not included as separate items in the motor score, but patients who do not walk are assigned to the most dependent response category for the walking items.

RTI calculated weights for the motor items using ordinary least squares regression to estimate the relative importance of each motor item in predicting wage-adjusted costs of care. The model included each motor item (or pair as noted above), as well as age at admission, the communication items sum score (Section BB), and the BIMS (Section C). The coefficients for each item were divided by the mean wage-adjusted cost in the sample to determine their relative size. These values represented the initial weights. The weights were then rescaled to a weighted average of 1 across each of the motor items so the weighted and unweighted motor scores would have the same range. Note that RTI found a high degree of correlation in the weighted and unweighted motor scores (Appendix D), but use of the weighted motor score was associated with a small increase the predictive power of the overall model (R-squared 0.3341 unweighted versus 0.3359 weighted; adjusted R-square 0.3338 unweighted versus 0.3355 weighted).

2.4 Cognitive Function Using Standardized Patient Assessment Data Elements

Standardized data elements related to cognitive function are collected in Section B and Section C of the IRF-PAI. Section B contains two items pertaining to communication: item BB0700 measures expression of ideas and wants, and Item BB0800 measures understanding verbal content. Section C contains an item measuring memory, the Brief Interview for Mental Status (BIMS). Though the IRF PPS includes a single cognitive score in the current CMG structure, the data elements in Section B and Section C cannot easily be summed. For this reason, two separate variables were used as independent variables in the CART model: first, a sum score of the two communication items in Section B, and second, memory using the BIMS in Section C.

Each of the communication items in Section B is scored on a 4-point scale, with higher scores indicating greater function. These two items were summed to a single communication score for inclusion in the CART model.



Section C of the IRF-PAI assesses memory using the BIMS. The BIMS consists of interview items and responses that sum to an overall score of 0 to 15, with a higher score indicating better memory skills. If the patient cannot complete the BIMS, a Staff Assessment of Mental Status is completed assessing memory. A three-level scale for memory using BIMS or the Staff Assessment of Mental Status was used in the CART model, as described in Table 8.

Table 8. Memory Using BIMS and Staff Assessment of Mental Status

Memory Category BIMS Score Staff Assessment Details Recalled

1 0–7 0–1

2 8–12 2

3 13–15 3–4

2.5 Costs

Costs of care are defined in this analysis as wage-adjusted costs for the IRF stay. IRF costs were calculated in three steps using data from the Medicare claims (Coomer, Ingber, Coots, & Morley, 2017). First, routine costs for the claim were calculated by multiplying the facility-specific routine cost per day (derived from the Cost Report) by the number of utilization days (length of stay) on the claim. The routine costs per day are based on the most up-to-date and complete cost report data available at the time of our analyses, FY 2017. These costs were inflated to reflect 2018 dollars using the inflation factors from the Inpatient Rehabilitation Facility Prospective Payment System for Federal Fiscal Year 2018 Final Rule. Claim ancillary costs were calculated by multiplying a set of 14 ancillary cost to charge ratios (derived from the facility-specific Cost Report) by the claim charges for those cost centers and summing. Ancillary costs for 2017 claims were inflated to reflect 2018 dollars; ancillary costs for 2018 claims were calculated with charges from 2018 and therefore not inflated. Finally, total claim cost was calculated as the sum of routine and ancillary costs. Total claim cost was then capped at the 0.5th and 99.9th percentile for freestanding IRFs and at the 0.2nd and 99.9th percentiles for IRF units. Wage-adjusted costs were calculated as follows:

Wage-Adjusted Cost = Claim Cost / (0.293 + 0.707 * Wage Index)

where 0.707 is the FY 2018 labor share of total costs of care (Centers for Medicare & Medicaid Services [CMS], 2017).

2.6 Results of CART Analysis

For each of the 21 RICs, RTI conducted CART analysis using the dependent variable of wage-adjusted cost of care for the stay, and four independent variables: (1) motor score,



(2) communication (IRF-PAI Section B), (3) memory (IRF-PAI Section C BIMS), and (4) beneficiary age on the date of admission. CART analysis was conducted using R statistical software.

The CART approach continues to split the sample until the contribution to model fit of any further splitting falls below a user-determined threshold. RTI set CART parameters to require the following:

• Nodes are no smaller than 100 stays.

• Additional splits yield an 0.5 percentage point increase in explanatory power (R-squared).

Results of initial regression tree runs for each RIC were reviewed to confirm monotonicity (i.e., that the level of functional limitation and average cost associated with each node increased across nodes within RIC) and consistency with clinical judgement. The R-squared statistics for the regression trees by node are shown in Appendix E.

The CART models using the standardized patient assessment data elements collected on admission resulted in 97 CMGs, including RICs 50 (“short stay,” one CMG) and 51 (“expired,” four CMGs), which remained unchanged. The CMGs based on the standardized patient assessment data elements collected on admission and age ranges used to group beneficiaries are presented in Table 9. The next section discusses calculation of the payment weights. For reference, the 92 CMGs in the FY 2019 IRF PPS are shown in Appendix B.

RTI adjusted the CART-generated trees for two RICs: RIC 12, “osteoarthritis,” and RIC 16, “pain syndrome.” A similar issue was observed in each of these RICs; the communication items emerged as a splitting variable in the CART analysis, but the thresholds for the splits were very high and not in line with clinical expectations. Specifically, the communication item was distinguishing beneficiaries with no impairment from all others (splitting at 6.5 and 7.5 out of 8) and attributing considerably higher cost for a lower level of impairment. Because of the very high threshold for this split, the inconsistency with clinical expectations, and the low number of observations in this RIC, the team removed the splits from the final CMG definitions.

Although the overall number of CMGs is similar to the FY 2019 CMGs using the standardized patient assessment data elements, the number of CMGs per RIC sometimes shifted. For example, RIC 1 contains 10 CMGs in the FY 2019 IRF PPS, but 7 when using the standardized patient assessment data elements. Each of the final CMGs is monotonic within each RIC. For a model using the CMGs based on the standardized patient assessment data elements and comorbidity tiers to predict wage-adjusted costs of care, the r-squared value



is 0.3358; the r-squared value is 0.3169 for the CMGs used in the current IRF PPS. In a model that also controls for the variation between providers, this value increases to 0.5409, compared with 0.5482 using the CMGs from the current IRF PPS. The r-squared value is higher in models controlling for variation between providers. Overall, there is little difference between the model r-squares using the CMGs from the current IRF PPS and the model r-squares using the CMGs based on the standardized patient assessment data elements. The statistics generated from the CART runs for each CMG are shown in Appendix E.

Motor score is the key function variable emerging in the definition of the CMGs across all RICs. Though included in the CART modeling, neither the communication items (IRF-PAI Section B) nor the memory items (IRF-PAI Section C BIMS) are reflected in the final CMG definitions. However, cognitive score is included in the CMG definitions for RIC 1, “stroke, and RIC 2, “traumatic brain injury,” in the FY 2019 IRF PPS. Though cognitive status is considered an important factor in resource use, current cognitive status items may not sufficiently measure the complexity of cognitive status, which may contribute to these results. Even without the explicit use of cognitive items in the CMG definitions, the function rating scale for the standardized patient assessment data elements may capture aspects of cognitive status; the scale measures need for assistance, including supervision.

Analyses to Inform

the Potential Use of S

tandardized Patient Assessm

ent Data

Elements in the Inpatient R

ehabilitation Facility Prospective Payment S

ystem

2-9

The findings and conclusions of this report are those of the authors and do not necessarily represent the view

s of HH

S.

Table 9. CMG Definitions Using Standardized Patient Assessment Data Elements

RIC CMG

CMG Description Relative Weight

Rule 1 Rule 2 Rule 3 Tier 1 Tier 2 Tier 3 None

Stroke 0101 Motor ≥ 72.00 1.0619 0.9248 0.8562 0.8152

Stroke 0102 Motor ≥ 63.90 Motor < 72.00 1.3354 1.1631 1.0768 1.0253




Stroke 0106 Motor < 40.90 Age ≥ 84.50 2.4326 2.1186 1.9615 1.8676

Stroke 0107 Motor < 40.90 Age < 84.50 2.8402 2.4736 2.2902 2.1805

Traumatic brain injury 0201 Motor ≥ 65.20 1.3159 1.0824 0.9892 0.9214

Traumatic brain injury 0202 Motor ≥ 55.05 Motor < 65.20 1.6232 1.3351 1.2201 1.1365

Traumatic brain injury 0203 Moto ≥ 49.90 Motor < 55.05 1.8426 1.5156 1.3851 1.2902

Traumatic brain injury 0204 Motor ≥ 34.65 Motor < 49.90 2.1349 1.7560 1.6048 1.4949

Traumatic brain injury 0205 Motor < 34.65 2.6896 2.2123 2.0218 1.8832

Non-traumatic brain injury 0301 Motor ≥ 69.20 1.1831 0.9602 0.8920 0.8326

Non-traumatic brain injury 0302 Motor ≥ 54.40 Motor < 69.20 1.5158 1.2303 1.1428 1.0668

Non-traumatic brain injury 0303 Motor ≥ 44.65 Motor < 54.40 1.8380 1.4917 1.3857 1.2935

Non-traumatic brain injury 0304 Motor < 44.65 Age ≥ 78.50 2.0873 1.6941 1.5737 1.4689

Non-traumatic brain injury 0305 Motor < 44.65 Age < 78.50 2.2569 1.8317 1.7015 1.5883

Traumatic spinal cord injury 0401 Motor ≥ 59.15 1.3469 1.1477 1.0636 0.9766

Traumatic spinal cord injury 0402 Moto ≥ 46.35 Motor < 59.15 1.8182 1.5493 1.4358 1.3184

Traumatic spinal cord injury 0403 Motor ≥ 38.10 Motor < 46.35 2.4146 2.0575 1.9067 1.7508

Traumatic spinal cord injury 0404 Motor < 32.45 Age < 61.50 3.1660 2.6978 2.5001 2.2956

Traumatic spinal cord injury 0405 Motor ≥ 32.45 Motor < 38.10 2.8545 2.4323 2.2541 2.0697

Traumatic spinal cord injury 0406 Motor ≥ 25.65 Motor < 32.45 Age ≥ 61.50 3.2618 2.7794 2.5757 2.3651

Traumatic spinal cord injury 0407 Motor < 25.65 Age ≥ 61.50 4.0436 3.4456 3.1931 2.9319

(continued)

Analyses to Inform



ent Data



ystem

2-1

0

The findings and conclusions of this report are those of the authors and do not necessarily represent the views of H

HS.

Table 9. CMG Definitions Using Standardized Patient Assessment Data Elements (continued)

RIC CMG



Non-traumatic spinal cord injury 0501 Motor ≥ 60.70 1.3019 1.0564 0.9906 0.9048

Non-traumatic spinal cord injury 0502 Motor ≥ 48.90 Motor < 60.70 1.7346 1.4075 1.3198 1.2055

Non-traumatic spinal cord injury 0503 Motor ≥ 40.40 Motor < 48.90 2.2683 1.8406 1.7259 1.5764

Non-traumatic spinal cord injury 0504 Motor < 40.40 2.8297 2.2961 2.1530 1.9666

Neurological 0601 Motor ≥ 66.60 1.3267 1.0265 0.9678 0.8781

Neurological 0602 Motor ≥ 53.90 Motor < 66.60 1.6480 1.2750 1.2022 1.0908

Neurological 0603 Motor ≥ 44.50 Motor < 53.90 1.9518 1.5101 1.4238 1.2918

Neurological 0604 Motor < 44.50 2.2464 1.7380 1.6387 1.4868

Fracture of lower extremity 0701 Motor ≥ 62.65 1.2794 1.0312 0.9863 0.8968

Fracture of lower extremity 0702 Motor ≥ 52.50 Motor < 62.65 1.6238 1.3089 1.2519 1.1383

Fracture of lower extremity 0703 Motor ≥ 44.00 Motor < 52.50 1.9191 1.5469 1.4795 1.3452

Fracture of lower extremity 0704 Motor < 44.00 2.1286 1.7157 1.6410 1.4921

Replacement of lower-extremity joint 0801 Motor ≥ 69.00 1.0169 0.8507 0.7719 0.7148

Replacement of lower-extremity joint 0802 Motor ≥ 56.80 Motor < 69.00 1.2485 1.0444 0.9477 0.8776

Replacement of lower-extremity joint 0803 Motor ≥ 45.45 Motor < 56.80 1.5244 1.2752 1.1571 1.0716

Replacement of lower-extremity joint 0804 Motor < 45.45 1.8673 1.5621 1.4175 1.3127

Other orthopedic 0901 Motor ≥ 64.95 1.2142 0.9706 0.9040 0.8322

Other orthopedic 0902 Motor ≥ 52.70 Motor < 64.95 1.5326 1.2251 1.1411 1.0504

Other orthopedic 0903 Motor ≥ 44.50 Motor < 52.70 1.8104 1.4471 1.3479 1.2408

Other orthopedic 0904 Motor < 44.50 2.0421 1.6324 1.5204 1.3996

Amputation lower extremity 1001 Motor ≥ 64.00 1.3062 1.1101 1.0101 0.9273

Amputation lower extremity 1002 Motor ≥ 51.90 Motor < 64.00 1.6752 1.4237 1.2954 1.1893

Amputation lower extremity 1003 Motor ≥ 46.00 Motor < 51.90 1.9319 1.6419 1.4939 1.3716

Amputation lower extremity 1004 Motor < 46.00 2.1597 1.8354 1.6701 1.5332

(continued)

Analyses to Inform



ent Data



ystem

2-1

1


HS.


RIC CMG



Amputation non-lower extremity 1101 Motor ≥ 58.60 1.4170 1.1613 1.0781 0.9074

Amputation non-lower extremity 1102 Motor ≥ 51.05 Motor < 58.60 1.8127 1.4856 1.3792 1.1608

Amputation non-lower extremity 1103 Motor < 51.05 2.0274 1.6616 1.5426 1.2983

Osteoarthritis 1201 Motor ≥ 59.45 1.3177 1.0136 0.9807 0.9023

Osteoarthritis 1202 Motor ≥ 49.90 Motor < 59.45 Age ≥ 81.50 1.6088 1.2376 1.1974 1.1017

Osteoarthritis 1203 Motor ≥ 49.90 Motor < 59.45 Age < 81.50 1.6351 1.2578 1.2170 1.1197

Osteoarthritis 1204 Motor < 49.90 1.8585 1.4297 1.3833 1.2727

Rheumatoid other arthritis 1301 Motor ≥ 64.35 1.1632 0.9757 0.9217 0.8541

Rheumatoid other arthritis 1302 Motor ≥ 49.45 Motor < 64.35 1.4774 1.2394 1.1708 1.0848

Rheumatoid other arthritis 1303 Motor < 49.45 Age ≥ 73.50 1.8461 1.5486 1.4629 1.3555

Rheumatoid other arthritis 1304 Motor < 49.45 Age < 73.50 1.9350 1.6232 1.5334 1.4208

Cardiac 1401 Motor ≥ 68.80 1.1626 0.9450 0.8778 0.7879

Cardiac 1402 Motor ≥ 59.10 Motor < 68.80 1.4251 1.1584 1.0760 0.9658

Cardiac 1403 Motor ≥ 48.60 Motor < 59.10 1.6815 1.3668 1.2696 1.1396

Cardiac 1404 Motor < 48.60 1.9763 1.6065 1.4922 1.3394

Pulmonary 1501 Motor ≥ 69.70 1.2419 1.0543 0.9813 0.9318

Pulmonary 1502 Motor ≥ 57.15 Motor < 69.70 1.5077 1.2799 1.1913 1.1312

Pulmonary 1503 Motor ≥ 44.60 Motor < 57.15 1.7841 1.5145 1.4096 1.3386

Pulmonary 1504 Motor < 44.60 2.0487 1.7391 1.6187 1.5371

Pain syndrome 1601 Motor ≥ 65.55 1.1679 0.9313 0.8775 0.8092

Pain syndrome 1602 Motor ≥ 56.65 Motor < 65.55 1.4665 1.1694 1.1019 1.0160

Pain syndrome 1603 Motor < 56.65 Age ≥ 71.50 1.7158 1.3682 1.2893 1.1888

Pain syndrome 1604 Motor < 56.65 Age < 71.50 1.7564 1.4006 1.3197 1.2169

(continued)

Analyses to Inform



ent Data



ystem

2-1

2


HS.


RIC CMG



Major multiple trauma without brain or spinal cord injury

1701 Motor ≥ 59.70

1.3943 1.0931 1.0271 0.9379


1702 Motor ≥ 47.00 Motor < 59.70

1.8097 1.4187 1.3331 1.2173


1703 Motor ≥ 37.80 Motor < 47.00

2.1547 1.6892 1.5872 1.4494


1704 Motor < 37.80

2.3848 1.8696 1.7567 1.6042

Major multiple trauma with brain or spinal cord injury

1801 Motor ≥ 71.60

1.0749 0.9247 0.8435 0.7703


1802 Motor ≥ 56.30 Motor < 71.60

1.4822 1.2751 1.1632 1.0623


1803 Motor ≥ 43.40 Motor < 56.30

1.9134 1.6460 1.5015 1.3712


1804 Motor ≥ 38.55 Motor < 43.40

2.2702 1.9530 1.7815 1.6270


1805 Motor ≥ 30.30 Motor < 38.55

2.6189 2.2530 2.0552 1.8769


1806 Motor < 30.30

3.4786 2.9925 2.7299 2.4930

Guillain-Barré 1901 Motor ≥ 60.85 1.2923 1.0458 1.0194 0.9800

Guillain-Barré 1902 Motor ≥ 49.80 Motor < 60.85 1.8782 1.5199 1.4816 1.4244

Guillain-Barré 1903 Motor ≥ 40.80 Motor < 49.80 2.5312 2.0483 1.9967 1.9196

Guillain-Barré 1904 Motor < 40.80 3.5306 2.8571 2.7850 2.6775

(continued)

Analyses to Inform



ent Data



ystem

2-1

3


HS.


RIC CMG



Miscellaneous 2001 Motor ≥ 65.95 1.2374 1.0001 0.9368 0.8491

Miscellaneous 2002 Motor ≥ 55.30 Motor < 65.95 1.5236 1.2315 1.1535 1.0455

Miscellaneous 2003 Motor ≥ 46.80 Motor < 55.30 1.7648 1.4264 1.3361 1.2110

Miscellaneous 2004 Motor < 46.80 Age ≥ 78.50 1.9471 1.5737 1.4740 1.3360

Miscellaneous 2005 Motor < 46.80 Age < 78.50 2.0925 1.6912 1.5841 1.4358

Burns 2101 Motor ≥ 53.90 1.5396 1.2552 1.1924 1.0556

Burns 2102 Motor < 53.90 2.1835 1.7802 1.6912 1.4970

Short stay 5001 0.1815

Mortality (orthopedic) LOS ≤ 13 5101 0.5698

Mortality (orthopedic) LOS ≥ 14 5102 1.7898

Mortality (non-orthopedic) LOS ≤ 15 5103 0.6737

Mortality (non-orthopedic) LOS ≥ 16 5104 2.1977

Note: LOS = length of stay.

Analyses to Inform the Potential Use of Standardized Patient Assessment Data Elements in the Inpatient Rehabilitation Facility Prospective Payment System


2.7 Payment Weight Calculations for Standardized Patient Data Element–Based CMGs

After generating new CMGs using the standardized patient assessment data elements collected at admission, RTI calculated payment weights for these groups. Payment weights are calculated for each CMG and comorbidity tier combination. Standardized payment amounts can then be calculated by multiplying the standard payment conversion factor by the relative weight associated with the CMG and comorbidity tier combination. CMS recalibrates these weights on an annual basis using cost report data for IRF stays from the prior year. RTI implemented the same approach to calculating payment weights as CMS uses in the current IRF PPS. The sections below describe the process.

Average Length of Stay

To calculate payment weights, first, the average length of stay (LOS) for every combination of CMG and comorbidity tier was calculated. This is done using an iterative process, where outlier cases with a length of stay more than three standard deviations from the mean are trimmed after each iteration. This process is repeated over five iterations, at which point the average length of stay is stabilized.

Comorbidity-Adjusted Costs

The second step of this process involves removing the effect of comorbidities from the wage-adjusted cost of care for each IRF stay. The effect of comorbidities on cost was estimated using ordinary least squares regression with provider fixed effects. The dependent variable in this analysis was log-transformed, wage-adjusted costs for the IRF stay. The model controlled for RIC, CMG (nested within RIC), and an interaction between RIC and comorbidity tier that accounted for a differential effect of comorbidities across RICs. Short transfers (defined as a transfer with a length of stay less than the average for that CMG and tier) and cases with log-transformed costs of care more than three standard deviations from the overall mean were excluded to reduce the effect of extreme value cases in the model. The cost of care for each case assigned to a comorbidity tier was then divided by the exponentiated regression coefficient for the corresponding RIC-tier combination to estimate what the stay’s cost of care would have been without comorbidities (tier-adjusted).

Calculating CMG-Level Relative Weights

The third step of this process estimates the relative payment weight for every CMG using the tier-adjusted costs of care. These weights are calculated using the Hospital-Specific Relative Value (HSRV) methodology described in detail in earlier work (Carter et al., 2002). The HSRV methodology uses an iterative process to determine the relative costliness of the IRF stays assigned to a particular CMG while adjusting for the relative costliness of the providers who cared for those patients. For example, consider two patients assigned to the same CMG with the same total costs of care. If patient A was treated at an IRF with higher



average costs of care than patient B, patient A would have a lower relative cost under the HSRV approach. The case-mix index (CMI), or relative costliness of each provider, is determined by the average cost of care across all of the provider’s patients. Short transfers are included in this step and given reduced weight in the calculation.

The weight for each case is set to 1 for all cases except short transfers, where weight equals the following:

Case Weight = (LOS + 0.5)/Average LOS

The CMG-level payment weight is initially calculated as follows:

Pmt. Weight = sum(Cost of Care)/sum(Case Weight)

CMI is then calculated for each provider as a function of payment weights and case weights:

CMI = sum(Pmt. Weight)/sum(Case Weight)

The CMG-level payment weight for each stay is then multiplied by the CMI for the provider in which the stay occurred, which yields a new CMG-level average payment weight, and subsequently, a new value for the CMI for each provider. This process is repeated over five iterations, at which point the values for CMG-level payment weight and CMI stabilize. At this stage, the case-weighted average payment weight across all cases is equal to 1.

Calculating Comorbidity-Level Weights

The final step in this process is adjusting the newly calculated payment weight for the effect of comorbidities. Essentially, the process used to remove the effect of comorbidities in the earlier step is reversed. The CMG-level payment weights are multiplied by the exponentiated RIC-Tier–level regression coefficients, which were initially used to estimate the tier-adjusted costs of care, to calculate a payment weight for every combination of CMG and comorbidity tier. The same tier-level multiplier is applied to every CMG within a RIC. For example, the RIC-Tier multiplier for RIC 1 (“stroke”) Tier B is approximately 1.31. Therefore, the payment weights for each CMGs under RIC 1 are multiplied by 1.31 to generate payment weights for patients assigned to comorbidity Tier B in each CMG in RIC 1.

Budget Neutrality Adjustment

Because costs of care can vary from year to year, CMS typically adjusts the final payment weights to ensure budget neutrality across years. To make this adjustment, first, the case-weighted average payment weight is calculated across all cases for the new and current payment weights. Next, the new payment weights are multiplied by the ratio of the case-weighted average of the legacy weights to the new weights.



𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵𝑁𝑁𝑁𝑁𝑁𝑁 𝐹𝐹𝑁𝑁𝐹𝐹𝐵𝐵𝐹𝐹𝑁𝑁 = 𝑠𝑠𝐵𝐵𝑠𝑠(𝑃𝑃𝑠𝑠𝐵𝐵.𝑊𝑊𝐵𝐵𝑊𝑊𝐵𝐵ℎ𝐵𝐵𝑠𝑠 𝐶𝐶𝐵𝐵𝑁𝑁𝑁𝑁𝐵𝐵𝐶𝐶𝐵𝐵)/𝑠𝑠𝐵𝐵𝑠𝑠(𝐶𝐶𝑁𝑁𝑠𝑠𝐵𝐵 𝑊𝑊𝐵𝐵𝑊𝑊𝐵𝐵ℎ𝐵𝐵𝑠𝑠 𝐶𝐶𝐵𝐵𝑁𝑁𝑁𝑁𝐵𝐵𝐶𝐶𝐵𝐵)

𝑠𝑠𝐵𝐵𝑠𝑠(𝑃𝑃𝑠𝑠𝐵𝐵.𝑊𝑊𝐵𝐵𝑊𝑊𝐵𝐵ℎ𝐵𝐵𝑠𝑠 𝑁𝑁𝐵𝐵𝑁𝑁)/𝑠𝑠𝐵𝐵𝑠𝑠(𝐶𝐶𝑁𝑁𝑠𝑠𝐵𝐵 𝑊𝑊𝐵𝐵𝑊𝑊𝐵𝐵ℎ𝐵𝐵𝑠𝑠 𝑁𝑁𝐵𝐵𝑁𝑁)

𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵𝑁𝑁𝑁𝑁𝑁𝑁 𝑊𝑊𝐵𝐵𝑊𝑊𝐵𝐵ℎ𝐵𝐵 = 𝑃𝑃𝑠𝑠𝐵𝐵.𝑊𝑊𝐵𝐵𝑊𝑊𝐵𝐵ℎ𝐵𝐵 ∗ 𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵𝑁𝑁𝑁𝑁𝑁𝑁 𝐹𝐹𝑁𝑁𝐹𝐹𝐵𝐵𝐹𝐹𝑁𝑁

The final list of payment weights by CMG and comorbidity tier is presented in Table 9. Table 10 presents descriptive statistics of the budget-neutral payment weights using standardized patient assessment data elements and the FY19 IRF PPS payment weights at the RIC level. At the RIC level, the changes in mean weight are relatively small.

Analyses to Inform



ent Data



ystem

2-1

7


HS.

Table 10. Comparison of RIC-Level Average Payment Weights

RIC

Payment Weights Using Standardized Patient Assessment Data Elements FY 2019 IRF PPS Payment Weights

Obs. Mean Std. Dev. Minimum Maximum Mean

Std. Dev. Minimum Maximum

1 Stroke 1.565 0.499 0.8152 2.8402 1.559 0.508 0.6451 2.7655 150,349

2 Traumatic brain injury 1.381 0.389 0.9214 2.6896 1.377 0.391 0.5527 2.4863 24,264

3 Non-traumatic brain injury 1.321 0.318 0.8326 2.2569 1.327 0.320 0.8135 2.1203 54,219

4 Traumatic spinal cord injury 1.949 0.764 0.9766 4.0436 1.946 0.785 0.6855 3.6175 5,673

5 Non-traumatic spinal cord injury 1.465 0.467 0.9048 2.8297 1.456 0.448 0.6070 2.6996 30,152

6 Neurological 1.331 0.301 0.8781 2.2464 1.350 0.316 0.6948 2.2148 107,287

7 Fracture of lower extremity 1.341 0.256 0.8968 2.1286 1.333 0.251 0.7171 1.9907 77,032

8 Replacement of lower extremity joint 0.995 0.207 0.7148 1.8673 0.995 0.219 0.5754 1.8691 31,477

9 Other orthopedic 1.187 0.235 0.8322 2.0421 1.199 0.252 0.6894 2.0372 58,644

10 Amputation lower extremity 1.510 0.342 0.9273 2.1597 1.493 0.335 0.7584 2.0247 19,125

11 Amputation non-lower extremity 1.418 0.330 0.9074 2.0274 1.395 0.320 0.8832 1.9208 832

12 Osteoarthritis 1.151 0.188 0.9023 1.8585 1.173 0.196 0.7877 1.7067 1,566

13 Rheumatoid other arthritis 1.208 0.251 0.8541 1.9350 1.208 0.223 0.8342 1.7337 1,794

14 Cardiac 1.173 0.262 0.7879 1.9763 1.169 0.263 0.6103 1.8581 42,789

15 Pulmonary 1.261 0.244 0.9318 2.0487 1.258 0.260 0.7596 1.9395 14,992

16 Pain syndrome 1.091 0.192 0.8092 1.4006 1.099 0.212 0.7954 1.8637 2,302

17 Major multiple trauma without brain or spinal cord injury

1.369 0.283 0.9379 2.3848 1.391 0.298 0.8196 2.3097 13,884

18 Major multiple trauma with brain or spinal cord injury

1.606 0.580 0.7703 3.4786 1.608 0.498 0.7943 2.6145 3,727

(continued)

Analyses to Inform



ent Data



ystem

2-1

8


HS.

Table 10. Comparison of RIC-Level Average Payment Weights (continued)

RIC

Payment Weights Using Standardized Patient Assessment Data Elements FY 2019 IRF PPS Payment Weights

Obs. Mean Std. Dev. Minimum Maximum Mean

Std. Dev. Minimum Maximum

19 Guillain-Barré 1.971 0.773 0.9800 3.5306 1.951 0.827 0.9096 4.2669 1,479

20 Miscellaneous 1.239 0.273 0.8491 2.0925 1.241 0.276 0.6500 1.9734 92,936

21 Burns 1.450 0.328 1.0556 2.1835 1.468 0.167 1.3168 1.9075 440

50 Short stay 0.181 0.000 0.1815 0.160 0.000 0.1599 17,241

51 Mortality 0.833 0.477 0.5698 2.1977 0.947 0.403 0.7539 2.1145 1,225

Total 1.321 0.422 1.321 0.428 753,429

R-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

REFERENCES

Carter, G. M., Buntin, M. B., Hayden, O., Kawata, J., Paddock, S. M., Relles, D. A., … Wynn, B. O. (2002). Analyses for the initial implementation of the Inpatient Rehabilitation Facility Prospective Payment System. Prepared for the Centers for Medicare and Medicaid Services. Santa Monica, CA: RAND Health. Retrieved from https://www.rand.org/content/dam/rand/pubs/monograph_reports/MR1500/MR1500.pdf .

Centers for Medicare & Medicaid Services (CMS). (2017). Medicare Program; Inpatient Rehabilitation Facility Prospective Payment System for Federal Fiscal Year 2018. Federal Register, 82 (148), 36238-36305. Retrieved from https://www.gpo.gov/fdsys/pkg/FR-2017-08-03/pdf/2017-16291.pdf

Coomer, N. M., Ingber, M., Coots, L. A., & Morley, M. (2017). Using Medicare cost reports to calculate costs for post-acute care claims. (RTI Press Publication No. OP-0036-1701). Research Triangle Park, NC: RTI Press. http://dx.doi.org/10.3768/rtipress.2017.op.0036.1701

Morgan, J. (2014, May 8). Classification and regression tree analysis. Boston, MA: Boston University School of Public Health, Department of Health Policy & Management. Retrieved from https://www.bu.edu/sph/files/2014/05/MorganCART.pdf

Morley, M., Silver, B., Deutsch, A., & Ingber, M. (2018). Analyses to inform the potential use of standardized patient assessment elements in the Inpatient Rehabilitation Facility Prospective Payment System. Retrieved from https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/InpatientRehabFacPPS/Downloads/IRFPPSAnalysis2018RTI.pdf

https://www.rand.org/content/dam/rand/pubs/monograph_reports/MR1500/MR1500.pdf

https://www.rand.org/content/dam/rand/pubs/monograph_reports/MR1500/MR1500.pdf

https://www.gpo.gov/fdsys/pkg/FR-2017-08-03/pdf/2017-16291.pdf

http://dx.doi.org/10.3768/rtipress.2017.op.0036.1701

https://www.bu.edu/sph/files/2014/05/MorganCART.pdf

https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/InpatientRehabFacPPS/Downloads/IRFPPSAnalysis2018RTI.pdf

https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/InpatientRehabFacPPS/Downloads/IRFPPSAnalysis2018RTI.pdf


R-2 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.


A-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

APPENDIX A IRF PATIENT ASSESSMENT INSTRUMENT

DEPARTMENT OF HEALTH AND HUMAN SERVICES OMB No. 0938-0842 CENTER FOR MEDICARE & MEDICAID SERVICES

PRA Disclosure Statement*

According to the Paperwork Reduction Act of 1995, no persons are required to respond to a collection of information unless it displays a valid OMB control number. The valid OMB control number for this information collection is 0938-0842. The time required to complete this information collection is estimated to average 54.5 minutes per response, including the time to review instructions, search existing data resources, gather the data needed, and complete and review the information collection. If you have comments concerning the accuracy of the time estimate(s) or suggestions for improving this form, please write to: CMS, 7500 Security Boulevard, Attn: PRA Reports Clearance Officer, Mail Stop C4-26-05, Baltimore, Maryland 21244-1850.

*This statement applies to the 2015 release of the IRF-PAI (version 1.3) and not to any additional burden related to theaddition of new data elements added for the purpose of informating CMS' newly adopted measures, which were finalized through the FY 2016 IRF PPS Final Rule, including those quality measures related to the IMPACT Act of 2014.

CENTER FOR MEDICARE & MEDICAID SERVICES ____________________________________________________________________________________________________________________________________

DEPARTMENT OF HEALTH AND HUMAN SERVICES OMB No. 0938-0842

Identification Information* Payer Information*

1. Facility InformationA. Facility Name

_____________________________________________________________ _____________________________________________________________ _____________________________________________________________ _____________________________________________________________ _____________________________________________________________

B. Facility Medicare Provider Number ________________________

2. Patient Medicare Number ____________________________________3. Patient Medicaid Number ____________________________________4. Patient First Name _________________________________________5A. Patient Last Name _________________________________________ 5B. Patient Identification Number ________________________________ 6. Birth Date ____/____/_______

MM / DD / YYYY

20. Payment Source(02 - Medicare Fee For Service; 51- Medicare-Medicare Advantage;99 - Not Listed)

A. Primary Source _________ B. Secondary Source _________

Medical Information*

21. Impairment Group ________ ________ Admission Discharge

Condition requiring admission to rehabilitation; code according to Appendix A.

22. Etiologic Diagnosis A. ______ (Use ICD codes to indicate the etiologic problem B. ______ that led to the condition for which the patient is receiving C. _______ rehabilitation)

23. Date of Onset of Impairment ____/____/________ MM / DD / YYYY

24. Comorbid Conditions 7. Social Security Number _____________________________________8. Gender (1 - Male; 2 - Female) ________________________________9. Race/Ethnicity (Check all that apply)

American Indian or Alaska Native A. ________

Asian B. ________

Black or African American C. ________

Hispanic or Latino D. ________

Native Hawaiian or Other Pacific Islander E. ________

White F. ________

10. Marital Status ______________ (1 - Never Married; 2 - Married; 3 - Widowed;4 - Separated; 5 - Divorced)

11. Zip Code of Patient's Pre-Hospital Residence _______________ 12. Admission Date ____/____/_______

MM / DD / YYYY 13. Assessment Reference Date ____/____/_____

MM / DD / YYYY 14. Admission Class

(1 - Initial Rehab; 2 - Evaluation; 3 - Readmission;4 - Unplanned Discharge; 5 - Continuing Rehabilitation)

Use ICD codes to enter comorbid medical conditionsA. J. S. B. K. T. C. L. U. D. M. V. E. N. W. F. O. X. G. P. Y. H. Q. I. R.

24A. Are there any arthritis conditions recorded in items #21, #22, or #24 that meet all of the regulatory requirements for IRF classification (in 42 CFR 412.29(b)(2)(x), (xi), and (xii))? ______ __________

(0 - No; 1 - Yes)

25. DELETED26. DELETED

Height and Weight (While measuring if the number is X.1-X.4 round down, X.5 or greater round up)

15A. Admit From (01- Home (private home/apt., board/care, assisted living, group home, transitional living); 02- Short-term General Hospital; 03 - Skilled Nursing Facility (SNF); 04 - Intermediate care; 06 - Home under care of organized home health service organization; 50 - Hospice (home);51 - Hospice (institutional facility); 61 - Swing bed; 62 - Another Inpatient Rehabilitation Facility; 63 - Long-Term Care Hospital (LTCH);64 - Medicaid Nursing Facility; 65 - Inpatient Psychiatric Facility; 66 - Critical Access Hospital; 99 - Not Listed)

16A. Pre-hospital Living Setting _______________ Use codes from 15A. Admit From

17. Pre-hospital Living With ______________ (Code only if item 16A is 01- Home: Code using 01 - Alone; 02 - Family/Relatives; 03 - Friends; 04 - Attendant; 05 - Other)

18. DELETED19. DELETED

25A. Height on admission (in inches) _____________________________

26A. Weight on admission (in pounds) ____________________________ Measure weight consistently, according to standard facility practice (e.g., in

a.m. after voiding, with shoes off, etc.) 27. Swallowing Status ________ ________

Admission Discharge 3- Regular Food: solids and liquids swallowed safely without supervision or

modified food consistency 2- Modified Food Consistency/Supervision: subject requires modified food

consistency and/or needs supervision for safety 1- Tube/Parenteral Feeding: tube/parenteral feeding used wholly or partially

as a means of sustenance 28. DELETED

IRF-PAI Version 1.4 Effective October 1, 2016 Page 1

INPATIENT REHABILITATION FACILITY - PATIENT ASSESSMENT INSTRUMENT



Function Modifiers* 39. FIMTM Instrument*

Complete the following specific functional items prior to scoring the FIMTM Instrument:

Admission Discharge

29. Bladder Level of Assistance (Score using FIM Levels 1 - 7)

30. Bladder Frequency of Accidents (Score as below) 7 - No accidents 6 - No accidents; uses device such as a catheter 5 - One accident in the past 7 days 4 - Two accidents in the past 7 days3 - Three accidents in the past 7 days 2 - Four accidents in the past 7 days 1 - Five or more accidents in the past 7 days Enter in Item 39G (Bladder) the lower (more dependent) score from Items 29 and 30 above

Admission Discharge

31. Bowel Level of Assistance (Score using FIM Levels 1 - 7)

32. Bowel Frequency of Accidents (Score as below) 7 - No accidents 6 - No accidents; uses device such as a ostomy 5 - One accident in the past 7 days 4 - Two accidents in the past 7 days3 - Three accidents in the past 7 days 2 - Four accidents in the past 7 days 1 - Five or more accidents in the past 7 days Enter in Item 39H (Bowel) the lower (more dependent) score of Items 31and 32 above.

Admission Discharge

33. Tub Transfer 34. Shower Transfer

(Score Items 33 and 34 using FIM Levels 1 - 7; use 0 if activity does not occur) See training manual for scoring of Item 39K (Tub/Shower Transfer)

Admission Discharge

35. Distance Walked

36. Distance Traveled in Wheelchair (Code items 35 and 36 using: 3 - 150 feet; 2 - 50 to 149 feet; 1 - Less than 50 feet; 0 – activity does not occur)

Admission Discharge

37. Walk

38. Wheelchair (Score Items 37 and 38 using FIM Levels 1 - 7; 0 if activity does not occur) See training manual for scoring of Item 39L (Walk/Wheelchair)

* The FIM data set, measurement scale and impairment codes incorporated or referenced herein are the property of U B Foundation Activities, Inc. ©1993, 2001 U B Foundation Activities, Inc. The FIM mark is owned by UBFA, Inc.

Admission Discharge SELF-CARE

A. Eating B. Grooming C. Bathing D. Dressing - Upper E. Dressing - Lower F. Toileting

SPHINCTER CONTROL

G. Bladder H. Bowel

TRANSFERS

I. Bed, Chair, Wheelchair J. Toilet K. Tub, Shower

W - Walk C - Wheelchair

LOCOMOTION B - Both

L. Walk/Wheelchair M. Stairs

A - Auditory V - Visual

COMMUNICATION B - Both

N. Comprehension O. Expression

V - Vocal N - Nonvocal

B - Both

SOCIAL COGNITION

P. Social Interaction Q. Problem Solving R. Memory

Goal

FIM LEVELS No Helper

7 Complete Independence (Timely, Safely) 6 Modified Independence (Device)

Helper - Modified Dependence 5 Supervision (Subject = 100%) 4 Minimal Assistance (Subject = 75% or more) 3 Moderate Assistance (Subject = 50% or more)

Helper - Complete Dependence 2 Maximal Assistance (Subject = 25% or more) 1 Total Assistance (Subject less than 25%)

0 Activity does not occur; Use this code only at admission



______________


Discharge Information* Therapy Information

40. Discharge Date ____/____/________ MM / DD / YYYY

41. Patient discharged against medical advice? ______________

42. Program Interruption(s)

43. Program Interruption Dates (Code only if item 42 is 1 - Yes)

A. 1st Interruption Date

MM / DD / YYYY

C. 2nd Interruption Date

MM / DD / YYYY

E. 3rd Interruption Date

MM / DD / YYYY

(0 - No; 1 - Yes)

(0 - No; 1 - Yes)

B. 1st Return Date

MM / DD / YYYY

D. 2nd Return Date

MM / DD / YYYY

F. 3rd Return Date

MM / DD / YYYY

44C. Was the patient discharged alive? _____________ (0 - No; 1 - Yes)

44D. Patient’s discharge destination/living setting, using codes below: (answer only if 44C = 1; if 44C = 0, skip to item 46)

(01- Home (private home/apt., board/care, assisted living, group home, transitional living); 02- Short-term General Hospital; 03 - Skilled Nursing Facility (SNF); 04 - Intermediate care; 06 - Home under care of organized home health service organization; 50 - Hospice (home); 51 - Hospice (institutional facility); 61 - Swing bed; 62 - Another Inpatient Rehabilitation Facility; 63 - Long-Term Care Hospital (LTCH); 64 - Medicaid Nursing Facility; 65 - Inpatient Psychiatric Facility; 66 - Critical Access Hospital; 99 - Not Listed)

45. Discharge to Living With _____________ (Code only if item 44C is 1 - Yes and 44D is 01 - Home; Code using 1 -Alone; 2 - Family / Relatives; 3 - Friends; 4 - Attendant; 5 - Other)

46. Diagnosis for Interruption or Death _____________ (Code using ICD code)

47. Complications during rehabilitation stay (Use ICD codes to specify up to six conditions that began with this rehabilitation stay)

A. ____________ B. ____________ C. ____________ D. ____________ E. ____________ F. ____________

* The FIM data set, measurement scale and impairment codes incorporated or referenced herein are the property of U B Foundation Activities, Inc. © 1993, 2001 U B Foundation Activities, Inc. The FIM mark is owned by UBFA, Inc.

O0401. Week 1: Total Number of Minutes Provided O0401A: Physical Therapy a. Total minutes of individual therapy ________ b. Total minutes of concurrent therapy ________ c. Total minutes of group therapy _______ d. Total minutes of co-treatment therapy ________

O0401B: Occupational Therapy a. Total minutes of individual therapy ________ b. Total minutes of concurrent therapy _________ c. Total minutes of group therapy _______ d. Total minutes of co-treatment therapy ________

O0401C: Speech-Language Pathology a. Total minutes of individual therapy ________ b. Total minutes of concurrent therapy _________ c. Total minutes of group therapy _______ d. Total minutes of co-treatment therapy ________

O0402. Week 2: Total Number of Minutes Provided O0402A: Physical Therapy a. Total minutes of individual therapy ________ b. Total minutes of concurrent therapy _________ c. Total minutes of group therapy _______ d. Total minutes of co-treatment therapy ________

O0402B: Occupational Therapy a. Total minutes of individual therapy ________ b. Total minutes of concurrent therapy _________ c. Total minutes of group therapy _______ d. Total minutes of co-treatment therapy ________

O0402C: Speech-Language Pathology a. Total minutes of individual therapy ________ b. Total minutes of concurrent therapy _________ c. Total minutes of group therapy _______ d. Total minutes of co-treatment therapy ________


.

OMB No. 0938-0842

Patient Identifier Date

INPATIENT REHABILITATION FACILITY - PATIENT ASSESSMENT INSTRUMENTQUALITY INDICATORS

ADMISSION

Section B. Hearing, Speech, and Vision. BB0700. Expression of Ideas and Wants (3-day assessment period)

Enter Code Expression of Ideas and Wants (consider both verbal and non-verbal expression and excluding language barriers) 4. Expresses complex messages without difficulty and with speech that is clear and easy to understand 3. Exhibits some difficulty with expressing needs and ideas (e.g., some words or finishing thoughts) or speech is not clear 2. Frequently exhibits difficulty with expressing needs and ideas 1. Rarely/Never expresses self or speech is very difficult to understand

BB0800. Understanding Verbal Content (3-day assessment period) Enter Code Understanding Verbal Content (with hearing aid or device, if used and excluding language barriers)

4. Understands: Clear comprehension without cues or repetitions 3. Usually Understands: Understands most conversations, but misses some part/intent of message. Requires cues at times to understand 2. Sometimes Understands: Understands only basic conversations or simple, direct phrases. Frequently requires cues to understand 1. Rarely/Never Understands

Section C. Cognitive Patterns. C0100. Should Brief Interview for Mental Status (C0200-C0500) be conducted? (3-day assessment period) Attempt to conduct interview with all patients.

Enter Code 0. No (patient is rarely/never understood) Skip to C0900. Memory/Recall Ability 1. Yes Continue to C0200. Repetition of Three Words

Brief Interview for Mental Status (BIMS)

C0200. Repetition of Three Words

Enter Code

Ask patient: “I am going to say three words for you to remember. Please repeat the words after I have said all three. The words are: sock, blue and bed. Now tell me the three words.”

Number of words repeated by patient after first attempt: 3. Three 2. Two 1. One 0. None

After the patient's first attempt say: “I will repeat each of the three words with a cue and ask you about them later: sock, something to wear; blue, a color; bed, a piece of furniture.” You may repeat the words up to two more times.

IRF-PAI Version 1.4 Effective October 1, 2016Quality Indicators - Admission Page 4

t

. .

..

.

.

.

.

.

.

.

OMB No. 0938-0842

Patien Identifier Date

Section C. Cognitive Patterns. Brief Interview for Mental Status (BIMS) - Continued

C0300. Temporal Orientation: Year, Month, Day

Enter Code

A. Ask patient: “Please tell me what year it is right now.” Patient's answer is: 3. Correct 2. Missed by 1 year 1. Missed by 2 to 5 years 0. Missed by more than 5 years or no answer

Enter Code

B. Ask patient: “What month are we in right now?” Patient's answer is: 2. Accurate within 5 days 1. Missed by 6 days to 1 month 0. Missed by more than 1 month or no answer

Enter Code C. Ask patient: “What day of the week is today?” Patient's answer is: 1. Correct 0. Incorrect or no answer

C0400. Recall

Enter Code

Ask patient: “Let's go back to the first question. What were those three words that I asked you to repeat?” If unable to remember a word, give cue (i.e., something to wear; a color; a piece of furniture) for that word.

A. Recalls “sock?” 2. Yes, no cue required 1. Yes, after cueing ("something to wear") 0. No, could not recall

Enter Code B. Recalls “blue?” 2. Yes, no cue required 1. Yes, after cueing ("a color") 0. No, could not recall

Enter Code C. Recalls “bed?” 2. Yes, no cue required 1. Yes, after cueing ("a piece of furniture") 0. No, could not recall

C0500. BIMS Summary Score.

Enter Score Add scores for questions C0200-C0400 and fill in total score (00-15) Enter 99 if the patient was unable to complete the interview

C0600. Should the Staff Assessment for Mental Status (C0900) be Conducted?

Enter Code 0. No (patient was able to complete Brief Interview for Mental Status) Skip to GG0100. Prior Functioning: Everyday Activities 1. Yes (patient was unable to complete Brief Interview for Mental Status) Continue to C0900. Memory/Recall Ability

Staff Assessment for Mental Status.

Do not conduct if Brief Interview for Mental Status (C0200-C0500) was completed.

C0900. Memory/Recall Ability.

Check all that the patient was normally able to recall

A. Current season

B. Location of own room

C. Staff names and faces

E. That he or she is in a hospital/hospital unit

Z. None of the above were recalled

Page 5 IRF-PAI Version 1.4 Effective October 1, 2016Quality Indicators - Admission

t

.

.

.

.

.

.

.

.

OMB No. 0938-0842

Patien Identifier Date

Section GG. Functional Abilities and Goals . GG0100. Prior Functioning: Everyday Activities. Indicate the patient's usual ability with everyday activities prior to the current illness, exacerbation, or injury.

3. Independent - Patient completed the activities by him/herself, with or without an assistive device, with no assistance from a helper.

2. Needed Some Help - Patient needed partial assistance from another person to complete activities.

1. Dependent - A helper completed the activities for the patient.

8. Unknown 9. Not Applicable

Enter Codes in Boxes

A. Self-Care: Code the patient's need for assistance with bathing, dressing, using the toilet, or eating prior to the current illness, exacerbation, or injury.

B. Indoor Mobility (Ambulation): Code the patient's need for assistance with walking from room to room (with or without a device such as cane, crutch, or walker) prior to the current illness, exacerbation, or injury.

C. Stairs: Code the patient's need for assistance with internal or external stairs (with or without a device such as cane, crutch, or walker) prior to the current illness, exacerbation, or injury.

D. Functional Cognition: Code the patient's need for assistance with planning regular tasks, such as shopping or remembering to take medication prior to the current illness, exacerbation, or injury.

GG0110. Prior Device Use. Indicate devices and aids used by the patient prior to the current illness, exacerbation, or injury.

Check all that apply

A. Manual wheelchair

B. Motorized wheelchair or scooter

C. Mechanical lift

D. Walker

E. Orthotics/Prosthetics

Z. None of the above


OMB No. 0938-0842


Section GG. Functional Abilities and Goals GG0130. Self-Care (3-day assessment period)

Code the patient's usual performance at admission for each activity using the 6-point scale. If activity was not attempted at admission, code the reason. Code the patient's discharge goal(s) using the 6-point scale. Do not use codes 07, 09, or 88 to code discharge goal(s).

CODING: Safety and Quality of Performance - If helper assistance is required because patient's performance is unsafe or of poor quality, score according to amount of assistance provided.

Activities may be completed with or without assistive devices. 06. Independent - Patient completes the activity by him/herself with no assistance from a helper. 05. Setup or clean-up assistance - Helper SETS UP or CLEANS UP; patient completes activity. Helper assists only prior to or following the activity. 04. Supervision or touching assistance - Helper provides VERBAL CUES or TOUCHING/STEADYING assistance as patient completes activity.

Assistance may be provided throughout the activity or intermittently. 03. Partial/moderate assistance - Helper does LESS THAN HALF the effort. Helper lifts, holds or supports trunk or limbs, but provides less than

half the effort. 02. Substantial/maximal assistance - Helper does MORE THAN HALF the effort. Helper lifts or holds trunk or limbs and provides more than half

the effort. 01. Dependent - Helper does ALL of the effort. Patient does none of the effort to complete the activity. Or, the assistance of 2 or more helpers is

required for the patient to complete the activity.

If activity was not attempted, code reason: 07. Patient refused 09. Not applicable 88. Not attempted due to medical condition or safety concerns

1. Admission

Performance.

2. Discharge

Goal. Enter Codes in Boxes

A. Eating: The ability to use suitable utensils to bring food to the mouth and swallow food once the meal is presented on a table/tray. Includes modified food consistency.

B. Oral hygiene: The ability to use suitable items to clean teeth. [Dentures (if applicable): The ability to remove and replace dentures from and to the mouth, and manage equipment for soaking and rinsing them.]

C. Toileting hygiene: The ability to maintain perineal hygiene, adjust clothes before and after using the toilet, commode, bedpan or urinal. If managing an ostomy, include wiping the opening but not managing equipment.

E. Shower/bathe self: The ability to bathe self in shower or tub, including washing, rinsing, and drying self. Does not include transferring in/out of tub/shower.

F. Upper body dressing: The ability to put on and remove shirt or pajama top; includes buttoning, if applicable.

G. Lower body dressing: The ability to dress and undress below the waist, including fasteners; does not include footwear.

H. Putting on/taking off footwear: The ability to put on and take off socks and shoes or other footwear that is appropriate for safe mobility.


I

OMB No. 0938-0842

Patient dentifier Date

Section GG. Functional Abilities and Goals GG0170. Mobility (3-day assessment period)








If activity was not attempted, code reason: 07. Patient refused 09. Not applicable 88. Not attempted due to medical condition or safety concerns

1. Admission

Performance.

2. Discharge

Goal. Enter Codes in Boxes

A. Roll left and right: The ability to roll from lying on back to left and right side, and return to lying on back.

B. Sit to lying: The ability to move from sitting on side of bed to lying flat on the bed.

C. Lying to sitting on side of bed: The ability to safely move from lying on the back to sitting on the side of the bed with feet flat on the floor, and with no back support.

D. Sit to stand: The ability to safely come to a standing position from sitting in a chair or on the side of the bed.

E. Chair/bed-to-chair transfer: The ability to safely transfer to and from a bed to a chair (or wheelchair).

F. Toilet transfer: The ability to safely get on and off a toilet or commode.

G. Car transfer: The ability to transfer in and out of a car or van on the passenger side. Does not include the ability to open/close door or fasten seat belt.

H1. Does the patient walk? 0. No, and walking goal is not clinically indicated

1. No, and walking goal is clinically indicated

2. Yes Continue to GG0170I. Walk 10 feet

Code the patient's discharge goal(s) for items GG0170I, J, K, L, M, N, O, and P. For admission performance, skip to GG0170Q1. Does the patient use a wheelchair/scooter?

Skip to GG0170Q1. Does the patient use a wheelchair/scooter?

I. Walk 10 feet: Once standing, the ability to walk at least 10 feet in a room, corridor or similar space.

J. Walk 50 feet with two turns: Once standing, the ability to walk at least 50 feet and make two turns.

K. Walk 150 feet: Once standing, the ability to walk at least 150 feet in a corridor or similar space.


OMB No. 0938-0842


Section GG. Functional Abilities and Goals GG0170. Mobility (3-day assessment period) - Continued








If activity was not attempted, code the reason: 07. Patient refused 09. Not applicable 88. Not attempted due to medical condition or safety concerns

1. Admission

Performance

2. Discharge

Goal.


L. Walking 10 feet on uneven surfaces: The ability to walk 10 feet on uneven or sloping surfaces, such as grass or gravel.

M. 1 step (curb): The ability to step over a curb or up and down one step.

N. 4 steps: The ability to go up and down four steps with or without a rail.

O. 12 steps: The ability to go up and down 12 steps with or without a rail.

P. Picking up object: The ability to bend/stoop from a standing position to pick up a small object, such as a spoon, from the floor.

Q1. Does the patient use a wheelchair/scooter? 0. No Skip to H0350. Bladder Continence 1. Yes Continue to GG0170R. Wheel 50 feet with two turns

R. Wheel 50 feet with two turns: Once seated in wheelchair/scooter, the ability to wheel at least 50 feet and make two turns.

RR1. Indicate the type of wheelchair/scooter used. 1. Manual 2. Motorized

S. Wheel 150 feet: Once seated in wheelchair/scooter, the ability to wheel at least 150 feet in a corridor or similar space.

SS1. Indicate the type of wheelchair/scooter used. 1. Manual 2. Motorized


.

.

.

.

.

.

OMB No. 0938-0842


Section H. Bladder and Bowel. H0350. Bladder Continence (3-day assessment period)

Enter Code Bladder continence - Select the one category that best describes the patient.

0. Always continent (no documented incontinence) 1. Stress incontinence only 2. Incontinent less than daily (e.g., once or twice during the 3-day assessment period) 3. Incontinent daily (at least once a day) 4. Always incontinent 5. No urine output (e.g., renal failure) 9. Not applicable (e.g., indwelling catheter)

H0400. Bowel Continence (3-day assessment period)

Enter Code Bowel continence - Select the one category that best describes the patient.

0. Always continent 1. Occasionally incontinent (one episode of bowel incontinence) 2. Frequently incontinent (2 or more episodes of bowel incontinence, but at least one continent bowel movement) 3. Always incontinent (no episodes of continent bowel movements) 9. Not rated, patient had an ostomy or did not have a bowel movement for the entire 3 days

Section I. Active Diagnoses. Comorbidities and Co-existing Conditions


I0900. Peripheral Vascular Disease (PVD) or Peripheral Arterial Disease (PAD)

I2900. Diabetes Mellitus (DM) (e.g., diabetic retinopathy, nephropathy, and neuropathy)

I7900. None of the above

Section J. Health Conditions. J1750. History of Falls

Enter Code Has the patient had two or more falls in the past year or any fall with injury in the past year? 0. No 1. Yes 8. Unknown

J2000. Prior Surgery

Enter Code Did the patient have major surgery during the 100 days prior to admission? 0. No 1. Yes 8. Unknown

Section K. Swallowing/Nutritional Status. K0110. Swallowing/Nutritional Status (3-day assessment period) Indicate the patient's usual ability to swallow.


A. Regular food - Solids and liquids swallowed safely without supervision or modified food or liquid consistency.

B. Modified food consistency/supervision - Patient requires modified food or liquid consistency and/or needs supervision during eating for safety.

C. Tube/parenteral feeding - Tube/parenteral feeding used wholly or partially as a means of sustenance.


.

.

.

OMB No. 0938-0842


Section M. Skin Conditions.

Report based on highest stage of existing ulcer(s) at its worst; do not "reverse" stage.

M0210. Unhealed Pressure Ulcer(s)

Enter Code Does this patient have one or more unhealed pressure ulcer(s) at Stage 1 or higher? 0. No Skip to O0100. Special Treatments, Procedures, and Programs 1. Yes Continue to M0300. Current Number of Unhealed Pressure Ulcers at Each Stage

M0300. Current Number of Unhealed Pressure Ulcers at Each Stage

Enter Number A. Stage 1: Intact skin with non-blanchable redness of a localized area usually over a bony prominence. Darkly pigmented skin may not

have a visible blanching; in dark skin tones only it may appear with persistent blue or purple hues.

Number of Stage 1 pressure ulcers

Enter Number B. Stage 2: Partial thickness loss of dermis presenting as a shallow open ulcer with a red or pink wound bed, without slough. May also

present as an intact or open/ruptured blister.

1. Number of Stage 2 pressure ulcers

Enter Number C. Stage 3: Full thickness tissue loss. Subcutaneous fat may be visible but bone, tendon or muscle is not exposed. Slough may be

present but does not obscure the depth of tissue loss. May include undermining and tunneling.


Enter Number D. Stage 4: Full thickness tissue loss with exposed bone, tendon or muscle. Slough or eschar may be present on some parts of the

wound bed. Often includes undermining and tunneling.


Enter Number E. Unstageable - Non-removable dressing: Known but not stageable due to non-removable dressing/device

1. Number of unstageable pressure ulcers due to non-removable dressing/device

Enter Number F. Unstageable - Slough and/or eschar: Known but not stageable due to coverage of wound bed by slough and/or eschar

1. Number of unstageable pressure ulcers due to coverage of wound bed by slough and/or eschar

Enter Number G. Unstageable - Deep tissue injury: Suspected deep tissue injury in evolution

1. Number of unstageable pressure ulcers with suspected deep tissue injury in evolution

Section O. Special Treatments, Procedures, and Programs O0100. Special Treatments, Procedures, and Programs

Check if treatment applies at admission

N. Total Parenteral Nutrition


I

OMB No. 0938-0842

Patient dentifier Date

DISCHARGE

Section GG. Functional Abilities and Goals GG0130. Self-Care (3-day assessment period)

Code the patient's usual performance at discharge for each activity using the 6-point scale. If activity was not attempted at discharge, code the reason.








3. Discharge.

Performance Enter Codes in Boxes

A. Eating: The ability to use suitable utensils to bring food to the mouth and swallow food once the meal is presented on a table/tray. Includes modified food consistency.

B. Oral hygiene: The ability to use suitable items to clean teeth. [Dentures (if applicable): The ability to remove and replace dentures from and to the mouth, and manage equipment for soaking and rinsing them.]

C. Toileting hygiene: The ability to maintain perineal hygiene, adjust clothes before and after using the toilet, commode, bedpan or urinal. If managing an ostomy, include wiping the opening but not managing equipment.

E. Shower/bathe self: The ability to bathe self in shower or tub, including washing, rinsing, and drying self. Does not include transferring in/out of tub/shower.

F. Upper body dressing: The ability to put on and remove shirt or pajama top; includes buttoning, if applicable.

G. Lower body dressing: The ability to dress and undress below the waist, including fasteners; does not include footwear.

H. Putting on/taking off footwear: The ability to put on and take off socks and shoes or other footwear that is appropriate for safe mobility.

Page 12 IRF-PAI Version 1.4 Effective October 1, 2016 Quality Indicators - Discharge

OMB No. 0938-0842


Section GG. Functional Abilities and Goals GG0170. Mobility (3-day assessment period)









3. Discharge.


A. Roll left and right: The ability to roll from lying on back to left and right side, and return to lying on back.

B. Sit to lying: The ability to move from sitting on side of bed to lying flat on the bed.

C. Lying to sitting on side of bed: The ability to safely move from lying on the back to sitting on the side of the bed with feet flat on the floor, and with no back support.

D. Sit to stand: The ability to safely come to a standing position from sitting in a chair or on the side of the bed.

E. Chair/bed-to-chair transfer: The ability to safely transfer to and from a bed to a chair (or wheelchair).

F. Toilet transfer: The ability to safely get on and off a toilet or commode.

G. Car transfer: The ability to transfer in and out of a car or van on the passenger side. Does not include the ability to open/ close door or fasten seat belt.

H3. Does the patient walk? 0. No 2. Yes Continue to GG0170I. Walk 10 feet

Skip to GG0170Q3. Does the patient use a wheelchair/scooter?

I. Walk 10 feet: Once standing, the ability to walk at least 10 feet in a room, corridor or similar space

J. Walk 50 feet with two turns: Once standing, the ability to walk at least 50 feet and make two turns

K. Walk 150 feet: Once standing, the ability to walk at least 150 feet in a corridor or similar space


OMB No. 0938-0842


Section GG. Functional Abilities and Goals GG0170. Mobility (3-day assessment period) - Continued









3. Discharge.


L. Walking 10 feet on uneven surfaces: The ability to walk 10 feet on uneven or sloping surfaces, such as grass or gravel.

M. 1 step (curb): The ability to step over a curb or up and down one step.

N. 4 steps: The ability to go up and down four steps with or without a rail.

O. 12 steps: The ability to go up and down 12 steps with or without a rail.

P. Picking up object: The ability to bend/stoop from a standing position to pick up a small object, such as a spoon, from the floor.

Q3. Does the patient use a wheelchair/scooter? 0. No Skip to J1800. Any Falls Since Admission 1. Yes Continue to GG0170R. Wheel 50 feet with two turns

R. Wheel 50 feet with two turns: Once seated in wheelchair/scooter, the ability to wheel at least 50 feet and make two turns.

RR3. Indicate the type of wheelchair/scooter used. 1. Manual 2. Motorized

S. Wheel 150 feet: Once seated in wheelchair/scooter, the ability to wheel at least 150 feet in a corridor or similar space.

SS3. Indicate the type of wheelchair/scooter used. 1. Manual 2. Motorized


.

.

.

2. Number of these Stage 2 pressure ulcers that were present upon admission - enter how many were noted at the time of admission

.

.

.

OMB No. 0938-0842


Section J. Health Conditions. J1800. Any Falls Since Admission . Enter Code Has the patient had any falls since admission?

0. No Skip to M0210. Unhealed Pressure Ulcer(s) 1. Yes Continue to J1900. Number of Falls Since Admission

J1900. Number of Falls Since Admission.

CODING: 0. None 1. One 2. Two or more


A. No ino c

njury: No evidence of any injury is noted on physical assessment by the nurse or primary care clinician; omplaints of pain or injury by the patient; no change in the patient's behavior is noted after the fall

B. Injury (except major): Skin tears, abrasions, lacerations, superficial bruises, hematomas and sprains; or any fall-related injury that causes the patient to complain of pain

C. Major injury: Bone fractures, joint dislocations, closed head injuries with altered consciousness, subdural hematoma

Section M. Skin Conditions.

Report based on highest stage of existing ulcer(s) at its worst; do not "reverse" stage.

M0210. Unhealed Pressure Ulcer(s)

Enter Code Does this patient have one or more unhealed pressure ulcer(s) at Stage 1 or higher?

0. No Skip to M0900A. Healed Pressure Ulcer(s) 1. Yes Continue to M0300. Current Number of Unhealed Pressure Ulcers at Each Stage

M0300. Current Number of Unhealed Pressure Ulcers at Each Stage

Enter Number A. Stage 1: Intact skin with non-blanchable redness of a localized area usually over a bony prominence. Darkly pigmented skin may not

have a visible blanching; in dark skin tones only it may appear with persistent blue or purple hues.

Number of Stage 1 pressure ulcers

Enter Number

Enter Number

B. Stage 2: Partial thickness loss of dermis presenting as a shallow open ulcer with a red or pink wound bed, without slough. May also present as an intact or open/ruptured blister.

1. Number of Stage 2 pressure ulcers If 0 Skip to M0300C. Stage 3


Enter Number

Enter Number

C. Stage 3: Full thickness tissue loss. Subcutaneous fat may be visible but bone, tendon or muscle is not exposed. Slough may be present but does not obscure the depth of tissue loss. May include undermining and tunneling.

1. Number of Stage 3 pressure ulcers If 0 Skip to M0300D. Stage 4



.

..

.

.

.

.

.

.

.

OMB No. 0938-0842


Section M. Skin Conditions M0300. Current Number of Unhealed Pressure Ulcers at Each Stage - Continued

Enter Number

Enter Number

D. Stage 4: Full thickness tissue loss with exposed bone, tendon or muscle. Slough or eschar may be present on some parts of the wound bed. Often includes undermining and tunneling.


1. Number of Stage 4 pressure ulcers If 0 Skip to M0300E. Unstageable - Non-removable dressing

Enter Number

Enter Number

E. Unstageable - Non-removable dressing: Known but not stageable due to non-removable dressing/device

1. Number of unstageable pressure ulcers due to non-removable dressing/device If 0 Skip to M0300F. Unstageable - Slough and/or eschar

2. Number of these unstageable pressure ulcers that were present upon admission - enter how many were noted at the time of admission

Enter Number

Enter Number

1. Number of unstageable pressure ulcers due to coverage of wound bed by slough and/or eschar If 0 Skip to M0300G. Unstageable - Deep tissue injury

F. Unstageable - Slough and/or eschar: Known but not stageable due to coverage of wound bed by slough and/or eschar


Enter Number

Enter Number

1. Number of unstageable pressure ulcers with suspected deep tissue injury in evolution If 0 Skip to M0800. Worsening in Pressure Ulcer Status Since Admission


G. Unstageable - Deep tissue injury: Suspected deep tissue injury in evolution

M0800. Worsening in Pressure Ulcer Status Since Admission

Indicate the number of current pressure ulcers that were not present or were at a lesser stage on admission. If no current pressure ulcer at a given stage, enter 0. Enter Number

A. Stage 2

Enter Number

B. Stage 3

Enter Number

C. Stage 4

Enter Number

D. Unstageable - Non-removable dressing

Enter Number

E. Unstageable - Slough and/or eschar

Enter Number

F. Unstageable - Deep tissue injury


OMB No. 0938-0842


Section M. Skin Conditions M0900. Healed Pressure Ulcer(s)

Indicate the number of pressure ulcers that were: (a) present on Admission; and (b) have completely closed (resurfaced with epithelium) upon Discharge. If there are no healed pressure ulcers noted at a given stage, enter 0.

Enter Number A. Stage 1

Enter Number B. Stage 2

Enter Number C. Stage 3

Enter Number D. Stage 4

Section O. Special Treatments, Procedures, and Programs O0250. Influenza Vaccine - Refer to current version of IRF-PAI Training Manual for current influenza vaccination season and reporting period.

Enter Code A. Did the patient receive the influenza vaccine in this facility for this year's influenza vaccination season?

0. No Skip to O0250C. If influenza vaccine not received, state reason 1. Yes Continue to O0250B. Date influenza vaccine received

B. Date influenza vaccine received Complete date and skip to Z0400A. Signature of Persons Completing the Assessment

M M D D Y Y Y Y

Enter Code C. If influenza vaccine not received, state reason:

1. Patient not in this facility during this year's influenza vaccination season 2. Received outside of this facility 3. Not eligible - medical contraindication 4. Offered and declined 5. Not offered 6. Inability to obtain influenza vaccine due to a declared shortage 9. None of the above


DEPARTMENT OF HEALTH AND HUMAN SERVICES OMB No. 0938-0842 CENTER FOR MEDICARE & MEDICAID SERVICES

Item Z0400A. Signature of Persons Completing the Assessment*

I certify that the accompanying information accurately reflects patient assessment information for this patient and that I collected or coordinated collection of this information on the dates specified. To the best of my knowledge, this information was collected in accordance with applicable Medicare and Medicaid requirements. I understand that this information is used as a basis for ensuring that patients receive appropriate and quality care, and as a basis for payment from federal funds. I further understand that payment of such federal funds and continued participation in the government-funded health care programs is conditioned on the accuracy and truthfulness of this information, and that I may be personally subject to or may subject my organization to substantial criminal, civil, and/or administrative penalties for submitting false information.

Signature Title Date Information is Provided Time

A.

B.

C.

D.

E.

F.

G.

H.

I.

J.

K.

L.


B-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

APPENDIX B FY 2019 IRF PPS CMGS AND PAYMENT WEIGHTS

CMG CMG Description

(M = Motor, C = Cognitive, A = Age)

Relative Weight

Tier 1 Tier 2 Tier 3 None

0101 Stroke M > 51.05

0.8465 0.7365 0.6747 0.6451

0102 Stroke M > 44.45 and M < 51.05 and C > 18.5

1.0706 0.9315 0.8533 0.8159

0103 Stroke M > 44.45 and M < 51.05 and C < 18.5

1.2391 1.0781 0.9876 0.9443

0104 Stroke M > 38.85 and M < 44.45

1.2938 1.1257 1.0312 0.9860

0105 Stroke M > 34.25 and M < 38.85

1.4871 1.2938 1.1852 1.1333

0106 Stroke M > 30.05 and M < 34.25

1.6628 1.4467 1.3253 1.2673

0107 Stroke M > 26.15 and M < 30.05

1.8653 1.6229 1.4867 1.4216

0108 Stroke M < 26.15 and A > 84.5

2.3056 2.0060 1.8376 1.7572

0109 Stroke M > 22.35 and M < 26.15 and A < 84.5

2.0857 1.8147 1.6624 1.5896

0110 Stroke M < 22.35 and A < 84.5

2.7655 2.4060 2.2041 2.1076

0201 Traumatic brain injury M > 53.35 and C > 23.5

0.8235 0.6628 0.5922 0.5527

0202 Traumatic brain injury M > 44.25 and M < 53.35 and C > 23.5

1.1508 0.9263 0.8275 0.7724

0203 Traumatic brain injury M > 44.25 and C < 23.5

1.2723 1.0240 0.9149 0.8539

0204 Traumatic brain injury M > 40.65 and M < 44.25

1.3841 1.1141 0.9953 0.9290


1.6330 1.3143 1.1743 1.0960


1.9661 1.5825 1.4139 1.3196

0207 Traumatic brain injury M < 22.05

2.4863 2.0012 1.7879 1.6687

0301 Non-traumatic brain injury M > 41.05

1.1727 0.9483 0.8703 0.8135

0302 Non-traumatic brain injury M > 35.05 and M < 41.05

1.4347 1.1603 1.0648 0.9953

0303 Non-traumatic brain injury M > 26.15 and M < 35.05

1.6572 1.3402 1.2300 1.1496



CMG CMG Description


Relative Weight


0304 Non-traumatic brain injury M < 26.15

2.1203 1.7147 1.5737 1.4709

0401 Traumatic spinal cord injury M > 48.45

1.0040 0.8097 0.7490 0.6855

0402 Traumatic spinal cord injury M > 30.35 and M < 48.45

1.4873 1.1996 1.1096 1.0155

0403 Traumatic spinal cord injury M > 16.05 and M < 30.35

2.3688 1.9105 1.7673 1.6175

0404 Traumatic spinal cord injury M < 16.05 and A > 63.5

4.0377 3.2566 3.0125 2.7571

0405 Traumatic spinal cord injury M < 16.05 and A < 63.5

3.6175 2.9177 2.6989 2.4701

0501 Non-traumatic spinal cord injury M > 51.35

0.9171 0.7145 0.6605 0.6070

0502 Non-traumatic spinal cord injury M > 40.15 and M < 51.35

1.2182 0.9491 0.8774 0.8063


1.5156 1.1809 1.0916 1.0031


1.7426 1.3577 1.2551 1.1533


1.9957 1.5550 1.4374 1.3209

0506 Non-traumatic spinal cord injury M < 23.75

2.6996 2.1034 1.9443 1.7867

0601 Neurological M > 47.75

1.0736 0.8242 0.7624 0.6948

0602 Neurological M > 37.35 and M < 47.75

1.3920 1.0686 0.9884 0.9008

0603 Neurological M > 25.85 and M < 37.35

1.7124 1.3146 1.2159 1.1082

0604 Neurological M < 25.85

2.2148 1.7003 1.5727 1.4334

0701 Fracture of lower extremity M > 42.15

1.0280 0.8387 0.7948 0.7171

0702 Fracture of lower extremity M > 34.15 and M < 42.15

1.3083 1.0674 1.0115 0.9127

0703 Fracture of lower extremity M > 28.15 and M < 34.15

1.5600 1.2728 1.2062 1.0883

0704 Fracture of lower extremity M < 28.15

1.9907 1.6242 1.5392 1.3888

0801 Replacement of lower extremity joint M > 49.55

0.8391 0.6841 0.6185 0.5754



CMG CMG Description


Relative Weight


0802 Replacement of lower extremity joint M > 37.05 and M < 49.55

1.0766 0.8777 0.7936 0.7382

0803 Replacement of lower extremity joint M > 28.65 and M < 37.05 and A > 83.5

1.4123 1.1514 1.0410 0.9684

0804 Replacement of lower extremity joint M > 28.65 and M < 37.05 and A < 83.5

1.2727 1.0376 0.9381 0.8727

0805 Replacement of lower extremity joint M > 22.05 and M < 28.65

1.5169 1.2367 1.1181 1.0401

0806 Replacement of lower extremity joint M < 22.05

1.8691 1.5238 1.3777 1.2816

0901 Other orthopedic M > 44.75

1.0283 0.8073 0.7481 0.6894

0902 Other orthopedic M > 34.35 and M < 44.75

1.3030 1.0230 0.9479 0.8736

0903 Other orthopedic M > 24.15 and M < 34.35

1.6262 1.2768 1.1831 1.0903

0904 Other orthopedic M < 24.15

2.0372 1.5995 1.4821 1.3659

1001 Amputation, lower extremity M > 47.65

1.0941 0.9260 0.8226 0.7584

1002 Amputation, lower extremity M > 36.25 and M < 47.65

1.3984 1.1835 1.0513 0.9693

1003 Amputation, lower extremity M < 36.25

2.0247 1.7136 1.5222 1.4034

1101 Amputation, non-lower extremity M > 36.35

1.3618 1.0044 1.0044 0.8832

1102 Amputation, non-lower extremity M < 36.35

1.9208 1.4167 1.4167 1.2458

1201 Osteoarthritis M > 37.65

1.1125 0.9541 0.8710 0.7877

1202 Osteoarthritis M > 30.75 and M < 37.65

1.4092 1.2085 1.1032 0.9978

1203 Osteoarthritis M < 30.75

1.7067 1.4637 1.3361 1.2084

1301 Rheumatoid, other arthritis M > 36.35

1.0977 0.9523 0.8893 0.8342

1302 Rheumatoid, other arthritis M > 26.15 and M < 36.35

1.4355 1.2454 1.1630 1.0909

1303 Rheumatoid, other arthritis M < 26.15

1.7337 1.5041 1.4046 1.3175

1401 Cardiac M > 48.85

0.9226 0.7511 0.6772 0.6103



CMG CMG Description


Relative Weight


1402 Cardiac M > 38.55 and M < 48.85

1.2379 1.0079 0.9086 0.8189

1403 Cardiac M > 31.15 and M < 38.55

1.4752 1.2011 1.0828 0.9759

1404 Cardiac M < 31.15

1.8581 1.5129 1.3639 1.2292

1501 Pulmonary M > 49.25

1.0145 0.8753 0.7927 0.7596

1502 Pulmonary M > 39.05 and M < 49.25

1.2970 1.1191 1.0134 0.9711

1503 Pulmonary M > 29.15 and M < 39.05

1.5391 1.3280 1.2026 1.1524

1504 Pulmonary M < 29.15

1.9395 1.6735 1.5155 1.4522

1601 Pain syndrome M > 37.15

1.2123 0.9280 0.8814 0.7954

1602 Pain syndrome M > 26.75 and M < 37.15

1.5361 1.1758 1.1169 1.0079

1603 Pain syndrome M < 26.75

1.8637 1.4266 1.3551 1.2228

1701 Major multiple trauma without brain or spinal cord injury M > 39.25

1.2825 0.9724 0.9103 0.8196

1702 Major multiple trauma without brain or spinal cord injury M > 31.05 and M < 39.25

1.5510 1.1760 1.1009 0.9912

1703 Major multiple trauma without brain or spinal cord injury M > 25.55 and M < 31.05

1.8097 1.3722 1.2846 1.1565

1704 Major multiple trauma without brain or spinal cord injury M < 25.55

2.3097 1.7513 1.6395 1.4761

1801 Major multiple trauma with brain or spinal cord injury M > 40.85

1.1285 1.0063 0.8504 0.7943

1802 Major multiple trauma with brain or spinal cord injury M > 23.05 and M < 40.85

1.6639 1.4838 1.2539 1.1712

1803 Major multiple trauma with brain or spinal cord injury M < 23.05

2.6145 2.3315 1.9703 1.8403

1901 Guillain-Barré M > 35.95

1.4000 1.0049 0.9440 0.9096



CMG CMG Description


Relative Weight


1902 Guillain-Barré M > 18.05 and M < 35.95

2.4651 1.7694 1.6622 1.6017

1903 Guillain-Barré M < 18.05

4.2669 3.0627 2.8772 2.7725

2001 Miscellaneous M > 49.15

0.9693 0.7709 0.7160 0.6500

2002 Miscellaneous M > 38.75 and M < 49.15

1.2597 1.0018 0.9306 0.8448

2003 Miscellaneous M > 27.85 and M < 38.75

1.5484 1.2314 1.1438 1.0384

2004 Miscellaneous M < 27.85

1.9734 1.5695 1.4578 1.3234

2101 Burns M > 0

1.9075 1.5493 1.4963 1.3168

5001 Short-stay cases, length of stay is 3 days or fewer

0.1599

5101 Expired, orthopedic, length of stay is 13 days or fewer

0.7539

5102 Expired, orthopedic, length of stay is 14 days or more

1.6493

5103 Expired, not orthopedic, length of stay is 15 days or fewer

0.8091

5104 Expired, not orthopedic, length of stay is 16 days or more

2.1145

SOURCE: Inpatient Rehabilitation Facility Prospective Payment System for Fiscal Year 2019; Final Rule.




C-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

APPENDIX C STANDARDIZED PATIENT ASSESSMENT DATA ELEMENTS AND

FUNCTIONAL INDEPENDENCE MEASURE (FIM™) ITEM DESCRIPTIVES

Table C-1. Standardized Patient Assessment Data Elements: Motor Score—Section GG

Motor Score—Section GG (Range 19–110) Mean SD

Motor Score 53.10 14.11

Analyses to Inform



ent Data



ystem

C-2


s of HH

S.

Table C-2. Standardized Patient Assessment Data Elements: Motor Items—Section GG

Motor Items—Section GG 1 2 3 4 5 6 7 9 88 Missing

GG0130A1 Admission Performance Eating

20,882 15,096 35,351 105,306 336,551 219,023 3,336 3,706 14,004 174

GG0130B1 Admission Performance Oral Hygiene

23,644 26,418 76,331 195,678 353,861 45,630 14,764 2,973 13,895 235

GG0130C1 Admission Performance Toileting Hygiene

167,306 160,889 160,006 165,403 31,484 21,849 14,099 7,414 24,737 242

GG0130E1 Admission Performance Bathing

68,331 146,503 320,108 126,825 22,669 3,638 22,987 3,872 38,195 301

GG0130F1 Admission Performance Upper Body Dressing

52,673 102,077 213,826 165,023 187,442 11,256 6,458 3,960 10,491 223

GG0130G1 Admission Performance Lower Body Dressing

194,261 234,005 172,632 110,765 19,187 4,134 6,025 2,296 9,935 189

GG0130H1 Admission Performance Footwear

280,305 186,164 111,820 88,664 50,360 6,968 6,625 5,525 16,767 231

GG0170A1 Admission Performance Roll Left Right Code

37,635 89,566 214,930 259,028 30,235 80,483 5,451 2,860 33,030 211

GG0170B1 Admission Performance Sit to Lying Code

54,445 116,409 257,269 238,251 23,468 42,713 3,651 2,212 14,816 195

GG0170C1 Admission Performance Lying to Sit Code

53,291 126,969 264,393 229,044 21,938 40,155 3,299 1,707 12,441 192

GG0170D1 Admission Performance Sit to Stand Code

68,626 109,089 311,605 214,692 7,856 7,328 2,728 3,535 27,784 186

GG0170E1 Admission Performance Chair/Bed to Chair Transfer Code

96,094 129,894 314,762 185,692 6,605 4,430 2,305 785 12,682 180

(continued)

Analyses to Inform



ent Data



ystem

C-3


s of HH

S.

Table C-2. Standardized Patient Assessment Data Elements: Motor Items—Section GG (continued)

Motor Items—Section GG 1 2 3 4 5 6 7 9 88 Missing

GG0170F1 Admission Performance Toilet Transfer Code

89,826 113,921 269,997 182,774 8,958 6,154 14,747 9,624 57,162 266

GG0170G1 Admission Performance Car Transfer Code

17,364 26,081 104,438 70,152 2,229 2,336 13,843 57,812 458,756 418

GG0170I1 Admission Performance Walk 10 Feet Code

66,091 31,791 234,649 219,530 5,119 3,646 3,675 1,390 63,391 124,147

GG0170J1 Admission Performance Walk 50 Feet Code

35,669 10,320 140,858 171,497 4,300 2,736 7,109 5,744 251,011 124,185

GG0170K1 Admission Performance Walk 150 Feet Code

28,721 5,022 52,277 93,215 3,240 2,313 9,744 12,470 422,195 124,232

GG0170L1 Admission Performance Walk 10 Feet on Uneven Surface

17,936 6,257 73,272 70,215 1,555 1,593 6,996 14,477 436,854 124,274

GG0170M1 Admission Performance 1 Step Code

26,004 17,868 131,882 97,741 1,962 1,622 11,391 11,582 329,161 124,216

GG0170N1 Admission Performance 4 Steps Code

19,551 9,359 108,537 101,394 2,139 1,678 11,914 17,923 356,690 124,244

GG0170O1 Admission Performance 12 Steps Code

17,572 2,944 20,561 37,243 1,335 1,497 14,124 33,591 500,271 124,291

GG0170P1 Admission Performance Pick Up Object Code

31,753 24,811 58,483 69,991 3,787 4,931 6,581 13,478 415,333 124,281

Analyses to Inform



ent Data



ystem

C-4


s of HH

S.

Table C-3. Standardized Patient Assessment Data Elements: Bowel and Bladder—Section H

Bowel and Bladder—Section H 0 1 2 3 4 5 9 Missing

H0350 Urinary Continence Code

426,855 38,491 100,310 74,735 46,251 9,279 57,404 104

H0400 Bowel Continence Code 547,546 73,768 41,531 47,326 43,152 106



Table C-4. Standardized Patient Assessment Data Elements: BIMS (Memory)—Section C

BIMS (Memory)—Section C (Range 1–15) Mean SD Not

Completed Missing

C0500 Brief Interview for Mental Status (BIMS) 14.22 12.34 5,386 40,357

Table C-5. Standardized Patient Assessment Data Elements: BIMS (Memory) Staff Assessment—Section C

BIMS (Memory) Staff Assessment—Section C 0 1 Missing

C0900A Staff Assessment of Mental Status—Recalls Current Season Code

36,413 16,579 700,437

C0900B Staff Assessment of Mental Status—Recalls Location of Room Code

42,056 10,892 700,481

C0900C Staff Assessment of Mental Status—Recalls Staff Name Code

41,573 11,385 700,471

C0900E Staff Assessment of Mental Status—Recalls Hospital Code

28,185 24,843 700,401

C0900Z Staff Assessment of Mental Status—Recalls None of Above Code

27,983 25,056 700,390

Table C-6. Standardized Patient Assessment Data Elements: Communication—Section B

Communication—Section B 1 2 3 4 Missing

BB0700 Expression Idea Want Code: Admission 19,876 62,833 284,505 385,691 524

BB0800 Understands Other Code: Admission 12,130 70,055 294,171 376,549 524



Table C-7. FIM™ Items Used in IRF PPS: FIM™ Motor Score

FIM™ Motor Score (Range 12–84) Mean SD

Motor Score 28.35 10.46

Table C-8. FIM™ Items Used in IRF PPS: Motor Items—FIM™

Motor Items—FIM™ 0 1 2 3 4 5 6 7

39AA Self-Care—Eating: Admission

2,078 55,033 16,644 22,871 72,213 468,110 59,801 56,679

39BA Self-Care—Grooming: Admission

6,078 97,110 40,815 77,957 196,536 321,675 7,318 5,940

39CA Self-Care—Bathing: Admission

32,678 158,576 109,945 216,439 184,622 49,064 1,652 453

39DA Self-Care—Dressing Upper: Admission

13,669 163,105 77,341 113,135 191,340 189,949 2,627 2,263

39EA Self-Care—Dressing Lower: Admission

9,650 380,964 133,841 98,370 104,916 24,427 853 408

39FA Self-Care—Toileting: Admission

13,071 334,977 129,223 103,151 136,957 32,107 2,928 1,015

39GA Sphincter Control—Bladder: Admission

0 333,829 53,653 56,256 61,463 147,701 50,144 50,383

39HA Sphincter Control—Bowel: Admission

0 216,266 50,914 48,635 54,836 103,039 238,382 41,357

39IA Transfers—Bed 3,347 248,044 150,990 188,631 150,678 10,805 710 224

39JA Transfers—Toilet: Admission

32,347 188,505 139,968 179,035 194,405 17,475 1,427 267

39LA Locomotion—Walk/Wheelchair: Admission

48,404 386,578 194,925 10,556 82,617 26,389 3,595 365

39MA Locomotion—Stairs: Admission

396,892 134,731 172,674 3,409 34,762 9,854 985 122

Table C-9. FIM™ Items Used in IRF PPS: FIM™ Cognitive Score

FIM™ Cognitive Score (Range 5–35) Mean SD

Cognitive Score 21.87 6.89



Table C-10. FIM™ Items Used in IRF PPS: Cognitive Items—FIM™

Cognitive Items—FIM™ 0 1 2 3 4 5 6 7

39NA Communication—Comprehension: Admission

0 23,136 50,978 98,690 154,278 218,118 161,404 46,825

39OA Communication—Expression: Admission

0 28,871 49,548 88,480 139,394 210,460 160,410 76,266

39PA Social Cognition—Social Interaction: Admission

0 23,188 39,791 77,302 127,890 219,800 185,887 79,571

39QA Social Cognition—Problem Solving: Admission

0 63,077 89,052 141,293 177,209 182,356 74,546 25,896

39RA Social Cognition—Memory: Admission

0 52,391 90,174 134,972 163,918 179,623 96,201 36,150




D-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

APPENDIX D CORRELATION OF WEIGHTED VERSUS UNWEIGHTED MOTOR

SCORE

Figure D-1. Weighted vs. Unweighted Motor Score


D-2 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.


E-1 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.

APPENDIX E CROSS VALIDATED R-SQUARED BY RIC

Table E-1. Cross Validated R-Squared from CART Models by RIC

RIC RIC Description

Split Number

1 2 3 4 5 6

1 Stroke 0.234 0.270 0.291 0.300 0.303 0.308

2 Traumatic brain injury 0.155 0.176 0.208 0.211

3 Non-traumatic brain injury 0.129 0.149 0.166 0.172

4 Traumatic spinal cord injury 0.223 0.253 0.272 0.293 0.297 0.308

5 Non-traumatic spinal cord injury 0.206 0.236 0.253

6 Neurological 0.114 0.131 0.149

7 Fracture of lower extremity 0.106 0.127 0.139

8 Replacement of lower extremity 0.101 0.128 0.143

9 Other orthopedic 0.101 0.122 0.134

10 Amputation, lower extremity 0.091 0.120 0.125

11 Amputation, non-lower extremity 0.045 0.062 1.000

12 Osteoarthritis 0.060 0.097 0.094 0.110 0.115

13 Rheumatoid, other arthritis 0.133 0.141 0.148

14 Cardiac 0.095 0.112 0.125

15 Pulmonary 0.093 0.108 0.125

16 Pain syndrome 0.060 0.078 0.092 0.088

17 Major multiple trauma without brain or spinal cord injury

0.109 0.145 0.152

18 Major multiple trauma with brain or spinal cord injury

0.202 0.225 0.275 0.278 0.291

19 Guillain-Barré 0.325 0.349 0.388

20 Miscellaneous 0.098 0.115 0.127 0.131

21 Burns −0.008 1.000 1.000


E-2 The findings and conclusions of this report are those of the authors and do not necessarily represent the views of HHS.


Documents

Analyses to Inform the Potential Use of Standardized ......Beginning in October 2016, standardized patient assessment data elements that measure functional status were introduced to